You are on page 1of 250

Mathematical Induction 101

With 101 Practice Problems

Duc Van Khanh Tran

Illustrations by Vy Nguyen Tong


Table of Contents

About the Author 1

Preface 2

Prerequisites: Mathematical Induction 4

Problems 1-10 7

Problems 11-20 28

Problems 21-30 49

Problems 31-40 71

Problems 41-50 93

Problems 51-60 118

Problems 61-70 140

Problems 71-80 164

Problems 81-90 190

Problems 91-101 216

Acknowledgements 245

Reviews 246
About the Author

Hi! My name is Duc Van Khanh Tran, and I am a Vietnamese undergraduate at


the University of Texas at Austin majoring in Mathematics.

When I was in elementary school, I studied at Morinosato Elementary School in


Kanazawa, Ishikawa province, Japan for about two years. For middle school edu-
cation, I studied at Le Quy Don Middle School in Ho Chi Minh City, Vietnam.

When I was in 9th grade, I came to the USA and started my high school study at
Brentwood Christian School in Austin, Texas. The classes in my high school were
more relaxing and less stressful than the classes in my middle school, so I began
understanding and enjoying the things I studied at school. After a while, I realized
that I enjoyed math the most out of all the subjects. Also, there was a math team
in my high school, and participating in the math team made me love mathematics
even more. That is how I “fell in love” with mathematics.

When I was in 11th grade, I came across @daily math , a very popular math account
on Instagram. Inspired by the @daily math page, I also created a math account
on Instagram called @dvkt math with about 33,000 followers currently. Other than
sharing mathematics knowledge by the means of a math page on Instagram, I de-
cided to also write math books. Before writing this book, I also wrote three other
books:
• An Introduction to Calculus: With Hyperbolic Functions, Limits, Derivatives,
and More (May 2021);
• Integrals and Sums Fiesta: An Integral Part of a Math Enthusiast’s Life
(November 2021);
• Basic Linear Algebra: An Introduction with an Intuitive Approach (May
2022).

1
Preface

Mathematical induction is one of the common methods of proof in mathematics.


It is a method to prove a statement about natural numbers n by figuring out the
pattern from some particular cases, such as n = 1, n = 2, n = 3, etc. The method
of mathematical induction is explained further in the “Prerequisites” chapter at
the beginning of the book.

After the “Prerequisites” chapter, there are 101 problems for the readers to practice
with mathematical induction. This book aims to introduce the method of math-
ematical induction to those who are unfamiliar with it while also providing some
interesting mathematical induction problems to those who are already familiar with
it. So, this book would be suitable for students who want to learn about mathemat-
ical induction or who are looking for more problems to practice with mathematical
induction.

Most of these problems are either created by myself or collected from online sources
such as Math Stack Exchange, Instagram, Twitter, etc. A few problems are taken
from or inspired by problems in the book The Cauchy-Schwarz Master Class by J.
Michael Steele and past preliminary exams of Singaporean junior colleges. (All of
the solutions are written by myself and not taken from others.)

While the practice problems cover a wide variety of topics (as you can see at the top
of each problem), most problems in this book only require the prerequisite knowl-
edge of high school Algebra and introductory Calculus. A few problems also require
some familiarity with the basics of Linear Algebra and Set Theory. There might
be some unfamiliar concepts to those who have only taken high school Algebra and
introductory Calculus, but some important definitions required to understand the
problems are provided in each problem.

Lastly, please note that since we are only proving using mathematical induction
in this book and the solutions are written by myself alone, some problems in this
book could have other ways to prove, either with or without using mathematical
induction. The readers are encouraged to find other possible methods of proof to
improve their proof skills.

I hope you enjoy the variety of mathematical induction problems awaiting ahead!
Duc Van Khanh Tran
Texas, USA, 2023

2
Prerequisites
Mathematical Induction

Mathematical induction is a proof method that can be used to prove a statement


about natural numbers n ∈ N. To clarify the notations, in this book N does not
include 0, and the letters k, m, and n always represent a natural number.

Before discussing the mathematical induction method itself, let us discuss why
it is called “mathematical induction.” There are mainly two types of proof: de-
ductive proof and inductive proof. A deductive proof is where we prove a general
statement directly by using axioms, definitions, proven theorems, etc. An inductive
proof is where we prove a general statement by proving some specific cases first and
then somehow relying on those specific cases to show that the statement is true
for all cases. As you will see, mathematical induction relies on some specific cases
called base cases.

In this book we will focus on two most common types of mathematical induc-
tion called weak induction and strong induction.

Weak Induction
To prove a statement S(n) for all natural numbers n, we prove the following.
1. S(1) is true;

2. If S(k) is true for some k ≥ 1, then S(k + 1) is also true.

Now let us discuss how this method can prove S(n) for all n ∈ N. First, we prove
that the base case S(1) is true. Then, we prove the induction step, which is S(k +1)
is true if S(k) is true. By applying the induction step to S(1), we can conclude
that S(2) is true. Then, we can also conclude that S(3) is true because S(2) is
true, S(4) is true because S(3) is true, S(5) is true because S(4) is true, and so on.
Thus, this method can prove that S(n) is true for all n ∈ N.

4
This method of weak induction can be modified slightly to prove a statement S(n)
for all n ≥ a for some natural number a > 1.

Weak Induction (Modified)


To prove a statement S(n) for all n ≥ a, we prove the following.

1. S(a) is true;
2. If S(k) is true for some k ≥ a, then S(k + 1) is also true.

Since we are proving a statement S(n) is true for all n ≥ a, the base case is S(a)
instead of S(1). Then, applying the induction step, we can conclude that S(a + 1),
S(a + 2), S(a + 3), etc. are true. Thus, we can prove that S(n) is true for all n ≥ a.

Strong Induction
To prove a statement S(n) for all natural numbers n, we prove the following.
1. S(1), S(2), · · · , S(`) are true for some natural number `;
2. If S(m) is true for all 1 ≤ m ≤ k for some k ≥ `, then S(k + 1) is also
true.

Now how does this method prove that S(n) is true for all n ∈ N? First, we prove
the base cases S(1), S(2), · · · , S(`). Then, we prove the induction step, which is
S(k + 1) is true if S(m) is true for all 1 ≤ m ≤ k. By applying the induction step
to S(1), S(2), · · · , S(`), we can conclude that S(` + 1) is true. Then, we can also
conclude that S(` + 2) is true because S(1), S(2), · · · , S(` + 1) are true, S(` + 3) is
true because S(1), S(2), · · · , S(` + 2) are true, and so on. Thus, this method can
prove that S(n) is true for all n ∈ N.

Similarly to weak induction, strong induction can also be modified slightly to prove
a statement S(n) for all n ≥ a for some natural number a > 1.

Strong Induction (Modified)


To prove a statement S(n) for all n ≥ a, we prove the following.
1. S(a), S(a + 1), · · · , S(a + `) are true for some natural number `;

2. If S(m) is true for all a ≤ m ≤ k for some k ≥ a + `, then S(k + 1) is


also true.

Since we are proving a statement S(n) is true for all n ≥ a, the base cases would
start with S(a) instead of S(1). So, the base cases would be S(a), S(a+1), · · · , S(a+

5
`). Then, applying the induction step, we can conclude that S(a+`+1), S(a+`+2),
S(a + ` + 3), etc. are true. Thus, we can prove that S(n) is true for all n ≥ a.

Now let us discuss the difference between weak induction and strong induction.
Note that in weak induction, we only need one previous case S(k) to prove S(k +1).
On the other hand, in strong induction, we need to use at least one previous case
other than S(k) to prove S(k + 1). Usually, we use two or more previous cases in
strong induction. However, like in problem 16, we do have some problems where
we only use one previous case in strong induction. In problem 16, we use S(k − 1),
NOT S(k), to prove S(k + 1), so it is strong induction instead of weak induction.
Thus, the difference between weak induction and strong induction mainly lies in
the induction step. That is why you will not be sure about whether to use weak or
strong induction and how many base cases you need to prove until you prove the
induction step.

When we prove the induction step, we need to assume that the previous case(s)
is/are true and then prove that S(k + 1) is true by using that assumption. The
induction step could be quite challenging because sometimes you would need to be
creative in connecting the previous cases to S(k + 1).

To be more familiar with mathematical induction and to be better at deciding


between weak and strong inductions, the best way is to practice as much as pos-
sible, and this book provides plenty of such practice problems. For each of the
problems in this book, I suggest that you try it on your own first and then look at
the solution after you figure it out or get stuck for a long time.

6
Problem 1
Topic: Series

Problem: Prove
n
X n(n + 1)
j= .
j=1
2

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1
X 1(1 + 1)
j=1= ,
j=1
2

so it is true for n = 1.

Induction Step
Assume
k
X k(k + 1)
j= (1.1)
j=1
2

for some k ≥ 1. To prove the induction step, we need to prove

k+1
X (k + 1)(k + 1 + 1) (k + 1)(k + 2)
j= = .
j=1
2 2

7
Using assumption (1.1),
k+1
X k
X
j= j + (k + 1)
j=1 j=1
k(k + 1)
= + (k + 1)
2
k(k + 1) 2(k + 1)
= +
2 2
(k + 1)(k + 2)
= .
2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X n(n + 1)
j=
j=1
2

for all n ∈ N.

8
Problem 2
Topic: Series

Problem: Prove
n
X n(n + 1)(2n + 1)
j2 = .
j=1
6

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1
X 1(1 + 1)(2 · 1 + 1)
j 2 = 12 = 1 = ,
j=1
6

so it is true for n = 1.

Induction Step
Assume
k
X k(k + 1)(2k + 1)
j2 = (2.1)
j=1
6

for some k ≥ 1. To prove the induction step, we need to prove

k+1
X (k + 1)(k + 1 + 1)(2(k + 1) + 1) (k + 1)(k + 2)(2k + 3)
j2 = = .
j=1
6 6

9
Using assumption (2.1),
k+1
X k
X
j2 = j 2 + (k + 1)2
j=1 j=1
k(k + 1)(2k + 1)
= + (k + 1)2
6
k(k + 1)(2k + 1) (k + 1)(6k + 6)
= +
6 6
(k + 1) [k(2k + 1) + (6k + 6)]
=
6 
(k + 1) 2k 2 + 7k + 6
=
6
(k + 1)(k + 2)(2k + 3)
= .
6
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X n(n + 1)(2n + 1)
j2 =
j=1
6

for all n ∈ N.

10
Problem 3
Topic: Series

Problem: Prove
n  2
X
3 n(n + 1)
j = .
j=1
2

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1  2
X 1(1 + 1)
j 3 = 13 = 1 = ,
j=1
2

so it is true for n = 1.

Induction Step
Assume
k  2
X
3 k(k + 1)
j = (3.1)
j=1
2

for some k ≥ 1. To prove the induction step, we need to prove

k+1  2  2
X
3 (k + 1)(k + 1 + 1) (k + 1)(k + 2)
j = = .
j=1
2 2

11
Using assumption (3.1),
k+1
X k
X
j3 = j 3 + (k + 1)3
j=1 j=1
 2
k(k + 1)
= + (k + 1)3
2
k 2 (k + 1)2 (4k + 4)(k + 1)2
= +
4 4
(k + 1)2 k 2 + 4k + 4
=
4
2
(k + 1)2 (k + 2)
=
4
 2
(k + 1)(k + 2)
= .
2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n  2
X
3n(n + 1)
j =
j=1
2

for all n ∈ N.

12
Problem 4
Topic: Harmonic Number, Inequality

Problem: Prove that Hn ≤ n for all n ∈ N where Hn denotes the n-th


Harmonic number.

Definition - Harmonic Number:


n
X 1
Hn = .
j=1
j

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
X 1 1
H1 = = = 1 ≤ 1,
j=1
j 1

so it is true for n = 1.

Induction Step
Assume
Hk ≤ k (4.1)
for some k ≥ 1. To prove the induction step, we need to prove

Hk+1 ≤ k + 1.

13
Using assumption (4.1),
k+1
X 1
Hk+1 =
j=1
j
k
X 1 1
= +
j=1
j k+1
1
= Hk +
k+1
1
≤k+ .
k+1
Since k + 1 ≥ 1,
1
≤ 1,
k+1
and we obtain
Hk+1 ≤ k + 1.
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

Hn ≤ n

for all n ∈ N.

14
Problem 5
Topic: Harmonic Number, Inequality

Problem: Prove
2n
Hn ≥
n+1
for all n ∈ N where Hn denotes the n-th Harmonic number.

Definition - Harmonic Number:


n
X 1
Hn = .
j=1
j

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
X 1 1 2·1
H1 = = =1≥ ,
j=1
j 1 1+1

so it is true for n = 1.

Induction Step
Assume
2k
Hk ≥ (5.1)
k+1

15
for some k ≥ 1. To prove the induction step, we need to prove

2(k + 1) 2k + 2
Hk+1 ≥ = .
(k + 1) + 1 k+2

Using assumption (5.1),


k+1
X 1
Hk+1 =
j=1
j
k
X 1 1
= +
j=1
j k+1
1
= Hk +
k+1
2k 1
≥ +
k+1 k+1
2k + 1
= .
k+1
Note that
2k + 1 2k + 2 k
− = ≥0
k+1 k+2 (k + 1)(k + 2)
because k ≥ 0, k + 1 > 0, and k + 2 > 0, so
2k + 1 2k + 2
≥ .
k+1 k+2
Therefore,
2k + 2
Hk+1 ≥ .
k+2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
2n
Hn ≥
n+1
for all n ∈ N.

16
Problem 6
Topic: Inequality

Problem: Prove en > n + 1 for all n ∈ N where e is the base of natural


logarithm, e ≈ 2.7183.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

e1 = e ≈ 2.7183 > 1 + 1,

so it is true for n = 1.

Induction Step
Assume
ek > k + 1 (6.1)
for some k ≥ 1. To prove the induction step, we need to prove

ek+1 > (k + 1) + 1 = k + 2.

Multiplying e to both sides of the inequality in assumption (6.1),

ek+1 > ek + e.

Since e > 2 > 1,


ek + e > k + 2.
Therefore,
ek+1 > k + 2.
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

17
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

en > n + 1

for all n ∈ N.

18
Problem 7
Topic: Inequality

Problem: Prove en−1 > 2n for n ≥ 4 where e is the base of natural logarithm,
e ≈ 2.7183.

Solution
Base Cases
We need to prove that it is true for n = 4. We can check that

e4−1 = e3 ≈ (2.7183)3 > (2.7)3 = 19.683 > 24 ,

so it is true for n = 4.

Induction Step
Assume
ek−1 > 2k (7.1)
for some k ≥ 4. To prove the induction step, we need to prove

e(k+1)−1 = ek > 2k+1 .

Multiplying e to both sides of the inequality in assumption (7.1),

ek > e · 2k .

Since e > 2,
e · 2k > 2 · 2k = 2k+1 .
Therefore,
ek > 2k+1 .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

19
Conclusion
We proved that it is true for n = 4. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

en−1 > 2n

for n ≥ 4.

20
Problem 8
Topic: Inequality

Problem: Prove n! > 2n + 2n for n ≥ 5.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 5. We can check that

5! = 120 > 42 = 2 · 5 + 25 ,

so it is true for n = 5.

Induction Step
Assume
k! > 2k + 2k (8.1)
for some k ≥ 5. To prove the induction step, we need to prove

(k + 1)! > 2(k + 1) + 2k+1 .

Multiplying both sides of assumption (8.1) by k + 1,

(k + 1)! > 2k(k + 1) + 2k (k + 1).

21
Since k ≥ 5 > 1 and k + 1 ≥ 6 > 2,

2k(k + 1) > 2(k + 1)

and
2k (k + 1) > 2k · 2 = 2k+1 .
So,
(k + 1)! > 2(k + 1) + 2k+1 .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 5. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

n! > 2n + 2n

for n ≥ 5.

22
Problem 9
Topic: Inequality

Problem: Prove nn > 3n2 + 6 for n ≥ 4.

Solution
Base Cases
We need to prove that it is true for n = 4. We can check that

44 > 3 · 42 + 6,

so it is true for n = 4.

Induction Step
Assume
k k > 3k 2 + 6 (9.1)
for some k ≥ 4. To prove the induction step, we need to prove

(k + 1)k+1 > 3(k + 1)2 + 6 = 3k 2 + 6k + 9.

Since k + 1 > k,

(k + 1)k+1 = (k + 1) · (k + 1)k > (k + 1) · k k .

Using assumption (9.1),

(k + 1) · k k > (k + 1) 3k 2 + 6 = 3k 3 + 3k 2 + 6k + 6.


Since k ≥ 4,
3k 3 + 6 ≥ 3 · 43 + 6 > 9,

23
so
3k 3 + 3k 2 + 6k + 6 > 3k 2 + 6k + 9.
Therefore,
(k + 1)k+1 > 3k 2 + 6k + 9 = 3(k + 1)2 + 6.
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 4. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

nn > 3n2 + 6

for n ≥ 4.

24
Problem 10
Topic: Inequality

Problem: Prove Bernoulli’s inequality, which states that

(1 + x)n ≥ 1 + nx

for x > −1 and n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

(1 + x)1 = 1 + x

and
1 + 1 · x = 1 + x,

it is indeed true that


(1 + x)1 ≥ 1 + 1 · x.

Induction Step
Assume
(1 + x)k ≥ 1 + kx (10.1)

for some k ≥ 1. To prove the induction step, we need to prove

(1 + x)k+1 ≥ 1 + (k + 1)x.

25
Since x > −1, 1 + x > 0. So, we can multiply 1 + x to both sides of the inequality
in assumption (10.1) without changing the sign:

(1 + x)k+1 ≥ (1 + x)(1 + kx)


= 1 + (k + 1)x + kx2 .

Since k > 0 and x2 ≥ 0, kx2 ≥ 0, and thus

(1 + x)k+1 ≥ 1 + (k + 1)x.

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

(1 + x)n ≥ 1 + nx

if x > −1 for all n ∈ N.

26
Problem 11
Topic: Series

Problem: For a 6= 1, prove


n−1
X an − 1
aj = .
j=0
a−1

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1−1
X a1 − 1
aj = a0 = 1 = ,
j=0
a−1

so it is true for n = 1.

Induction Step
Assume
k−1
X ak − 1
aj = (11.1)
j=0
a−1

for some k ≥ 1. To prove the induction step, we need to prove


(k+1)−1 k
X
j
X ak+1 − 1
a = aj = .
j=0 j=0
a−1

28
Using assumption (11.1),
k
X k−1
X
aj = aj + ak
j=0 j=0

ak − 1
= + ak
a−1
ak − 1 ak (a − 1)
= +
a−1 a−1
ak − 1 ak+1 − ak
= +
a−1 a−1
k+1
a −1
= .
a−1
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n−1
X an − 1
aj =
j=0
a−1

if a 6= 1 for all n ∈ N.

29
Problem 12
Topic: Series, Fibonacci Number

Problem: Prove
n
X
Fj = Fn+2 − 1
j=1

where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,

1
X
Fj = F1 = 1
j=1

and

F1+2 − 1 = F3 − 1 = F2 + F1 − 1 = F1 + F0 + F1 − 1 = 1 + 0 + 1 − 1 = 1,

so it is indeed true that


1
X
Fj = F1+2 − 1.
j=1

30
Induction Step
Assume
k
X
Fj = Fk+2 − 1 (12.1)
j=1

for some k ≥ 1. To prove the induction step, we need to prove


k+1
X
Fj = F(k+1)+2 − 1 = Fk+3 − 1.
j=1

Using assumption (12.1),


k+1
X k
X
Fj = Fj + Fk+1
j=1 j=1

= Fk+2 − 1 + Fk+1 .

Using the recursive relation of Fn ,

Fn = Fn−1 + Fn−2 ,

we have
Fk+2 + Fk+1 = Fk+3 .
Therefore,
k+1
X
Fj = Fk+3 − 1.
j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X
Fj = Fn+2 − 1
j=1

for all n ∈ N.

31
Problem 13
Topic: Series, Fibonacci Number

Problem: Prove
n
X
F2j+1 = F2n+2
j=0

where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,
1
X
F2j+1 = F1 + F3
j=0

and
F2·1+2 = F4 = F2 + F3 .
Since
F2 = F1 + F0 = F1 + 0 = F1 ,
it is indeed true that
1
X
F2j+1 = F2·1+2 .
j=0

32
Induction Step
Assume
k
X
F2j+1 = F2k+2 (13.1)
j=0

for some k ≥ 1. To prove the induction step, we need to prove


k+1
X
F2j+1 = F2(k+1)+2 = F2k+4 .
j=0

Using assumption (13.1),


k+1
X k
X
F2j+1 = F2j+1 + F2k+3
j=0 j=0

= F2k+2 + F2k+3 .

Using the recursive relation of Fn ,

Fn = Fn−1 + Fn−2 ,

we have
F2k+2 + F2k+3 = F2k+4 .
Therefore,
k+1
X
F2j+1 = F2k+4 .
j=0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X
F2j+1 = F2n+2
j=0

for all n ∈ N.

33
Problem 14
Topic: Series, Fibonacci Number

Problem: Prove
n
X
F2j = F2n+1 − 1
j=1

where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,

1
X
F2j = F2 = F1 + F0 = 1 + 0 = 1
j=1

and

F2·1+1 − 1 = F3 − 1 = F2 + F1 − 1 = F1 + F0 + F1 − 1 = 1 + 0 + 1 − 1 = 1,

so it is indeed true that


1
X
F2j = F2·1+1 − 1.
j=1

34
Induction Step
Assume
k
X
F2j = F2k+1 − 1 (14.1)
j=1

for some k ≥ 1. To prove the induction step, we need to prove


k+1
X
F2j = F2(k+1)+1 − 1 = F2k+3 − 1.
j=1

Using assumption (14.1),


k+1
X k
X
F2j = F2j + F2k+2
j=1 j=1

= F2k+1 − 1 + F2k+2 .

Using the recursive relation of Fn ,

Fn = Fn−1 + Fn−2 ,

we have
F2k+1 + F2k+2 = F2k+3 .
Therefore,
k+1
X
F2j = F2k+3 − 1.
j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X
F2j = F2n+1 − 1
j=1

for all n ∈ N.

35
Problem 15
Topic: Series, Fibonacci Number

Problem: Prove
n
X
Fj2 = Fn Fn+1
j=1

where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,

1
X
Fj2 = F12 = 1
j=1

and
F1 F1+1 = F1 F2 = F1 (F1 + F0 ) = 1,

so it is indeed true that


1
X
Fj2 = F1 F1+1 .
j=1

36
Induction Step
Assume
k
X
Fj2 = Fk Fk+1 (15.1)
j=1

for some k ≥ 1. To prove the induction step, we need to prove


k+1
X
Fj2 = Fk+1 F(k+1)+1 = Fk+1 Fk+2 .
j=1

Using assumption (15.1),


k+1
X k
X
Fj2 = Fj2 + Fk+1
2

j=1 j=1
2
= Fk Fk+1 + Fk+1
= Fk+1 (Fk + Fk+1 ).

Using the recursive relation of Fn ,

Fn = Fn−1 + Fn−2 ,

we have
Fk + Fk+1 = Fk+2 .
Therefore,
k+1
X
Fj2 = Fk+1 Fk+2 .
j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X
Fj2 = Fn Fn+1
j=1

for all n ∈ N.

37
Problem 16
Topic: Series

Problem: Prove
n
X 1 n+1
= n+1 .
j=0
2n + 4j 2

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. We can check that
1
X 1 1 1 1 1+1
= + = = 1+1
j=0
21 + 4 j 2+1 2+4 2 2

and
2
X 1 1 1 1 3 2+1
= + + = = 2+1 ,
j=0
22 + 4j 4 + 1 4 + 4 4 + 16 8 2
so it is true for n = 1 and n = 2.

Induction Step
Assume
m
X 1 m+1
= m+1 (16.1)
j=0
2m +4 j 2
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
k+1
X 1 (k + 1) + 1 k+2
= (k+1)+1 = k+2 .
j=0
2k+1 + 4j 2 2

38
Note that we can write the sum when n = k + 1 as
k+1 k
X 1 1 1 X 1
= + +
j=0
2k+1 + 4j 2k+1 + 1 2k+1 + 4k+1 j=1 2k+1 + 4j
k
1 1 1X 1
= k+1 + k+1 + .
2 +1 2 + 4k+1 4 j=1 2k−1 + 4j−1

Substituting j − 1 → j,
k k−1
X 1 X 1
k−1 j−1
= k−1 + 4j
.
j=1
2 +4 j=0
2

Then, by assumption (16.1) when m = k − 1,


k−1
X 1 (k − 1) + 1 k
= (k−1)+1 = k .
j=0
2k−1 + 4j 2 2

So,
k+1
X 1 1 1 1 k
= k+1 + + · k
j=0
2k+1 + 4j 2 + 1 2k+1 + 4k+1 4 2
1 1 k
= + + k+2
2k+1
+1 2k+1
(1 + 2 k+1 ) 2
k+1
2 +1 k
= k+1 k+1 +
2 (2 + 1) 2k+2
1 k
= k+1 + k+2
2 2
k+2
= k+2 .
2
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
n
X 1 n+1
= n+1
j=0
2n + 4 j 2

for all n ∈ N.

39
Problem 17
Topic: Divisibility

Problem: Prove
3 | n3 − n


for all n ∈ N.

Definition - Divisibility: For a ∈ Z \ {0} and b ∈ Z, a divides b, denoted as


a | b, when a is a factor of b. Formally,

a | b ⇔ b = a`

for some ` ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

13 − 1 = 0 = 3 · 0

where 0 is an integer,
3 | 13 − 1


by definition of divisibility.

Induction Step
Assume
3 | k3 − k

(17.1)

40
for some k ≥ 1. To prove the induction step, we need to prove

3 | (k + 1)3 − (k + 1) .


By assumption (17.1) and the definition of divisibility,

k 3 − k = 3`

for some ` ∈ Z. Then,

(k + 1)3 − (k + 1) = k 3 + 3k 2 + 3k + 1 − k − 1
= 3` + 3k 2 + 3k
= 3 ` + k2 + k .


Since ` + k 2 + k is an integer,

3 | (k + 1)3 − (k + 1)


by definition of divisibility. Thus, we proved that it is true for n = k + 1 if it is


true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

3 | n3 − n


for all n ∈ N.

41
Problem 18
Topic: Divisibility

Problem: Prove
5 | n5 − n


for all n ∈ N.

Definition - Divisibility: For a ∈ Z \ {0} and b ∈ Z, a divides b, denoted as


a | b, when a is a factor of b. Formally,

a | b ⇔ b = a`

for some ` ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

15 − 1 = 0 = 5 · 0

where 0 is an integer,
5 | 15 − 1


by definition of divisibility.

Induction Step
Assume
5 | k5 − k

(18.1)

42
for some k ≥ 1. To prove the induction step, we need to prove

5 | (k + 1)5 − (k + 1) .


By assumption (18.1) and the definition of divisibility,

k 5 − k = 5`

for some ` ∈ Z. Then,

(k + 1)5 − (k + 1) = k 5 + 5k 4 + 10k 3 + 10k 2 + 5k + 1 − k − 1


= 5` + 5k 4 + 10k 3 + 10k 2 + 5k
= 5 ` + k 4 + 2k 3 + 2k 2 + k .


Since ` + k 4 + 2k 3 + 2k 2 + k is an integer,

5 | (k + 1)5 − (k + 1)


by definition of divisibility. Thus, we proved that it is true for n = k + 1 if it is


true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

5 | n5 − n


for all n ∈ N.

43
Problem 19
Topic: Divisibility

Problem: Prove
7 | (9n − 2n )
for all n ∈ N.

Definition - Divisibility: For a ∈ Z \ {0} and b ∈ Z, a divides b, denoted as


a | b, when a is a factor of b. Formally,

a | b ⇔ b = a`

for some ` ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

91 − 21 = 7 = 7 · 1

where 1 is an integer,
7 | 91 − 21


by definition of divisibility.

Induction Step
Assume
7 | 9k − 2k

(19.1)

44
for some k ≥ 1. To prove the induction step, we need to prove

7 | 9k+1 − 2k+1 .


By assumption (19.1) and by definition of divisibility,

9k − 2k = 7`

for some ` ∈ Z. Then,

9k+1 − 2k+1 = 9 · 9k − 2 · 2k
= (7 + 2) · 9k − 2 · 2k
= 7 · 9k + 2 · 9k − 2k


= 7 · 9k + 2 · 7`
= 7 9k + 2` .


Since 9k + 2` is an integer,
7 | 9k+1 − 2k+1


by definition of divisibility. Thus, we proved that it is true for n = k + 1 if it is


true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

7 | (9n − 2n )

for all n ∈ N.

45
Problem 20
Topic: Divisibility

Problem: Prove
9 | 82n − 3 · 7n + 2


for all n ∈ N.

Definition - Divisibility: For a ∈ Z \ {0} and b ∈ Z, a divides b, denoted as


a | b, when a is a factor of b. Formally,

a | b ⇔ b = a`

for some ` ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

82·1 − 3 · 71 + 2 = 45 = 9 · 5

where 5 is an integer,
9 | 82·1 − 3 · 71 + 2


by definition of divisibility.

Induction Step
Assume
9 | 82k − 3 · 7k + 2

(20.1)

46
for some k ≥ 1. To prove the induction step, we need to prove
 
9 | 82(k+1) − 3 · 7k+1 + 2 .

By assumption (20.1) and by definition of divisibility,

82k − 3 · 7k + 2 = 9`

for some ` ∈ Z. Then,

82(k+1) − 3 · 7k+1 + 2 = 64 · 82k − 7 · 3 · 7k + 2


= 63 · 82k − 6 · 3 · 7k + 82k − 3 · 7k + 2
= 63 · 82k − 6 · 3 · 7k + 9`
= 9 7 · 82k − 2 · 7k + ` .


Since 7 · 82k − 2 · 7k + ` is an integer,


 
9 | 82(k+1) − 3 · 7k+1 + 2

by definition of divisibility. Thus, we proved that it is true for n = k + 1 if it is


true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

9 | 82n − 3 · 7n + 2


for all n ∈ N.

47
Problem 21
Topic: Inequality

Problem: For x, y ≥ 0, prove


n
xn + y n

x+y

2 2

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
x1 + y 1

x+y x+y
= ≤ ,
2 2 2

so it is true for n = 1.

Induction Step
Assume
k
xk + y k

x+y
≤ (21.1)
2 2
for some k ≥ 1. To prove the induction step, we need to prove
k+1
xk+1 + y k+1

x+y
≤ .
2 2

49
x+y
Multiplying both sides of the inequality in assumption (21.1) by 2 ,
k+1
xk + y k
   
x+y x+y

2 2 2
xk+1 + y k+1 + xy k + yxk
=
4
xk+1 + y k+1 xk+1 + y k+1 xy k + yxk
= − +
2 4 4
xk+1 + y k+1 xk+1 + y k+1 − xy k − yxk
= −
2  4
xk+1 + y k+1 xk − y k (x − y)
= − .
2 4
For any real numbers x and y, it is always either x ≥ y or y ≥ x. If x ≥ y, then
x − y ≥ 0 and xk − y k ≥ 0, so

xk − y k (x − y) ≥ 0.


If y ≥ x, then x − y ≤ 0 and xk − y k ≤ 0, so

xk − y k (x − y) ≥ 0.


In both cases,
xk − y k (x − y) ≥ 0.


So, 
xk − y k (x − y)
− ≤ 0,
4
and thus we have

xk+1 + y k+1 xk − y k (x − y) xk+1 + y k+1
− ≤ .
2 4 2
Therefore,
k+1
xk+1 + y k+1

x+y
≤ .
2 2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
xn + y n

x+y

2 2

for x, y ≥ 0 and for all n ∈ N.

50
Problem 22
Topic: Inequality

Problem: If a > b > 0, prove

(2a − 2b)n ≥ 2an − (2b)n

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
(2a − 2b)1 = 2a − 2b ≥ 2a1 − (2b)1 ,
so it is true for n = 1.

Induction Step
Assume
(2a − 2b)k ≥ 2ak − (2b)k (22.1)
for some k ≥ 1. To prove the induction step, we need to prove
(2a − 2b)k+1 ≥ 2ak+1 − (2b)k+1 .
Since a > b, 2a − 2b > 0, so we can multiply 2a − 2b to both sides of the inequality
in assumption (22.1) without changing the sign:
(2a − 2b)k+1 ≥ (2a − 2b) 2ak − (2b)k


= 4ak+1 + (2b)k+1 − 4bak − (2a)(2b)k


= 2ak+1 + 2ak+1 − (2b)k+1 + 2(2b)k+1 − 4bak − (2a)(2b)k
= 2ak+1 − (2b)k+1 + 2(a − 2b) ak − (2b)k .


51
For any real numbers a and 2b, it is always either a ≥ 2b or 2b ≥ a. If a ≥ 2b, then
a − 2b ≥ 0 and ak − (2b)k ≥ 0, so

(a − 2b) ak − (2b)k ≥ 0.


If 2b ≥ a, then a − 2b ≤ 0 and ak − (2b)k ≤ 0, so

(a − 2b) ak − (2b)k ≥ 0.


In both cases,
(a − 2b) ak − (2b)k ≥ 0.


So,
2ak+1 − (2b)k+1 + 2(a − 2b) ak − (2b)k ≥ 2ak+1 − (2b)k+1 .


Therefore,
(2a − 2b)k+1 ≥ 2ak+1 − (2b)k+1 .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

(2a − 2b)n ≥ 2an − (2b)n

if a > b > 0 for all n ∈ N.

52
Problem 23
Topic: Inequality

Problem: For y ≥ 0, prove

ny − y n ≤ n − 1

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. We can check that

1 · y − y1 = 0 ≤ 1 − 1

and
2 · y − y 2 = −(y − 1)2 + 1 ≤ 1 = 2 − 1,
so it is true for n = 1 and n = 2.

Induction Step
Assume
my − y m ≤ m − 1 (23.1)
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

(k + 1)y − y k+1 ≤ (k + 1) − 1 = k.

By assumption (23.1) when m = k,

ky − y k ≤ k − 1 ⇔ −y k ≤ k − 1 − ky.

53
Since y ≥ 0, we can multiply y to both sides of the inequality without changing the
sign:
−y k+1 ≤ y(k − 1 − ky).
So,

(k + 1)y − y k+1 ≤ (k + 1)y + y(k − 1 − ky)


= 2ky − ky 2
= k 2y − y 2 .


By assumption (23.1) when m = 2,

2y − y 2 ≤ 2 − 1 = 1.

Therefore,
(k + 1)y − y k+1 ≤ k.
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

ny − y n ≤ n − 1

if y ≥ 0 for all n ∈ N.

54
Problem 24
Topic: Matrices

Problem: For a, b ∈ R, prove


 n  n 
a 0 a 0
=
0 b 0 bn

for all n ∈ N.

Definition - Matrix Multiplication: For those who are unfamiliar with


matrices, for our purpose here, we only need to know that
    
a11 a12 b11 b12 a11 b11 + a12 b21 a11 b12 + a12 b22
= .
a21 a22 b21 b22 a21 b11 + a22 b21 a21 b12 + a22 b22

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
 1    1 
a 0 a 0 a 0
= = ,
0 b 0 b 0 b1
so it is true for n = 1.

Induction Step
Assume  k  k 
a 0 a 0
= (24.1)
0 b 0 bk

55
for some k ≥ 1. To prove the induction step, we need to prove
 k+1  k+1 
a 0 a 0
= .
0 b 0 bk+1

Using assumption (24.1),


 k+1  k  
a 0 a 0 a 0
=
0 b 0 b 0 b
 k  
a 0 a 0
=
0 bk 0 b
 k
a · a + 0 · 0 ak · 0 + 0 · b

=
0 · a + bk · 0 0 · 0 + bk · b
 k+1 
a 0
= .
0 bk+1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 n  n 
a 0 a 0
=
0 b 0 bn

for a, b ∈ R and for all n ∈ N.

56
Problem 25
Topic: Matrices

Problem: For a ∈ R, prove


 n  
1 a 1 na
=
0 1 0 1

for all n ∈ N.

Definition - Matrix Multiplication: For those who are unfamiliar with


matrices, for our purpose here, we only need to know that
    
a11 a12 b11 b12 a11 b11 + a12 b21 a11 b12 + a12 b22
= .
a21 a22 b21 b22 a21 b11 + a22 b21 a21 b12 + a22 b22

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
 1  
1 a 1 1·a
= ,
0 1 0 1
so it is true for n = 1.

Induction Step
Assume  k  
1 a 1 ka
= (25.1)
0 1 0 1

57
for some k ≥ 1. To prove the induction step, we need to prove
 k+1  
1 a 1 (k + 1)a
= .
0 1 0 1

Using assumption (25.1),


 k+1  k  
1 a 1 a 1 a
=
0 1 0 1 0 1
  
1 ka 1 a
=
0 1 0 1
 
1 · 1 + ka · 0 1 · a + ka · 1
=
0·1+1·0 0·a+1·1
 
1 (k + 1)a
= .
0 1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 n  
1 a 1 na
=
0 1 0 1

for all n ∈ N.

58
Problem 26
Topic: Matrices, Fibonacci Number

Problem: Prove  n  
1 1 F Fn
= n+1
1 0 Fn Fn−1
for all n ∈ N where Fn denotes the n-th Fibonacci number.

Definition - Matrix Multiplication: For those who are unfamiliar with


matrices, for our purpose here, we only need to know that
    
a11 a12 b11 b12 a11 b11 + a12 b21 a11 b12 + a12 b22
= .
a21 a22 b21 b22 a21 b11 + a22 b21 a21 b12 + a22 b22

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,
 1    
1 1 1 1 F F1
= = 2
1 0 1 0 F1 F0

because F0 = 0, F1 = 1, and F2 = F1 + F0 = 1 + 0 = 1, so it is true for n = 1.

59
Induction Step
Assume  k  
1 1 Fk+1 Fk
= (26.1)
1 0 Fk Fk−1
for some k ≥ 1. To prove the induction step, we need to prove
 k+1    
1 1 F(k+1)+1 Fk+1 Fk+2 Fk+1
= = .
1 0 Fk+1 F(k+1)−1 Fk+1 Fk

Using assumption (26.1),


 k+1  k  
1 1 1 1 1 1
=
1 0 1 0 1 0
  
Fk+1 Fk 1 1
=
Fk Fk−1 1 0
 
Fk+1 · 1 + Fk · 1 Fk+1 · 1 + Fk · 0
=
Fk · 1 + Fk−1 · 1 Fk · 1 + Fk−1 · 0
 
Fk + Fk+1 Fk+1
= .
Fk−1 + Fk Fk

Using the recursive relation

Fn = Fn−1 + Fn−2

of Fn ,
Fk + Fk+1 = Fk+2
and
Fk−1 + Fk = Fk+1 .
Therefore,
 k+1  
1 1 F Fk+1
= k+2 .
1 0 Fk+1 Fk
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 n  
1 1 Fn+1 Fn
=
1 0 Fn Fn−1

for all n ∈ N.

60
Problem 27
Topic: Floor Function, Inequality

Problem: Prove   j k
n+1 n
0≤ − ≤1
2 2
for all n ∈ N where b·c denotes the floor function.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. When n = 1,
   
1+1 1
− = 1 − 0 = 1,
2 2
so it is true that    
1+1 1
0≤ − ≤ 1.
2 2

61
Induction Step
Assume    
k+1 k
0≤ − ≤1 (27.1)
2 2
for some k ≥ 1. To prove the induction step, we need to prove
   
(k + 1) + 1 k+1
0≤ − ≤ 1.
2 2

Note that
       
(k + 1) + 1 k+1 k+2 k+1
− = −
2 2 2 2
   
k k+1
= +1 −
2 2
   
k k+1
= +1−
2 2
   
k+1 k
=1− − .
2 2

By assumption (27.1),    
k+1 k
0≤ − ≤ 1,
2 2
so    
k+1 k
0≤1− − ≤ 1.
2 2
Therefore,    
(k + 1) + 1 k+1
0≤ − ≤ 1.
2 2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
  j k
n+1 n
0≤ − ≤1
2 2

for all n ∈ N.

62
Problem 28
Topic: Floor Function, Inequality

Problem: Prove  
n+1 n

2 2
for all n ∈ N where b·c denotes the floor function.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. We can check that
 
1+1 1
=1≥
2 2
and  
2+1 2
=1≥ ,
2 2
so it is true for n = 1 and n = 2.

63
Induction Step
Assume  
m+1 m
≥ (28.1)
2 2
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
 
(k + 1) + 1 k+1
≥ .
2 2

Note that        
(k + 1) + 1 k+2 k k
= = +1 = + 1.
2 2 2 2
By assumption (28.1) when m = k − 1,
   
(k − 1) + 1 k k−1
= ≥ .
2 2 2

Therefore,  
(k + 1) + 1 k−1 k+1
≥ +1= .
2 2 2
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
 
n+1 n

2 2

for all n ∈ N.

64
Problem 29
Topic: Floor Function, Inequality

Problem: Prove √  n
n ≤
2
for n ≥ 2 where b·c denotes the floor function.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 2 and n = 3. When n = 2,
√ √ √ √
1 < 2 < 4 ⇔ 1 < 2 < 2,
so j√ k 2
2 =1≤ .
2
When n = 3, √ √ √ √
1< 3< 4⇔1< 3 < 2,

65
so j√ k 3
3 =1≤ .
2

Induction Step
Assume √  m
m ≤ (29.1)
2
for all 2 ≤ m ≤ k for some k ≥ 3. To prove the induction step, we need to prove
j√ k k+1
k+1 ≤ .
2
Note that for k ≥ 3, √ √
2 k − 1 ≥ 2 2 ≥ 1,
so √
k + 2 k − 1 ≥ k + 1.
Since
√ √ √ 2
k + 2 k − 1 = (k − 1) + 1 + 2 k − 1 = k−1+1 ,

we have the inequality


√ 2
k−1+1 ≥ k + 1.

Taking square root of both sides,


√ √
k − 1 + 1 ≥ k + 1.

So, j√ k j√ k j√ k
k+1 ≤ k−1+1 = k − 1 + 1.

By assumption (29.1) when m = k − 1,


j√ k k−1
k−1 ≤ .
2
Therefore,
j√ k k−1 k+1
k+1 ≤ +1= .
2 2
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 2 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 2 and n = 3. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 2 ≤ m ≤ k. Therefore, by strong induction,
√  n
n ≤
2
for n ≥ 2.

66
Problem 30
Topic: Floor Function

Problem: Prove a floor function identity of Ramanujan,


jnk n + 2 n + 4 jnk n + 3
+ + = + ,
3 6 6 2 6

for all n ∈ N where b·c denotes the floor function.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1, n = 2, n = 3, n = 4, n = 5, and n = 6.
We can check that
         
1 1+2 1+4 1 1+3
+ + = + = 0,
3 6 6 2 6

67
         
2 2+2 2+4 2 2+3
+ + = + = 1,
3 6 6 2 6

         
3 3+2 3+4 3 3+3
+ + = + = 2,
3 6 6 2 6

         
4 4+2 4+4 4 4+3
+ + = + = 3,
3 6 6 2 6

         
5 5+2 5+4 5 5+3
+ + = + = 3,
3 6 6 2 6

and
         
6 6+2 6+4 6 6+3
+ + = + = 4,
3 6 6 2 6

so it is true for n = 1, n = 2, n = 3, n = 4, n = 5, and n = 6.

Induction Step
Assume
jmk     j k  
m+2 m+4 m m+3
+ + = + (30.1)
3 6 6 2 6

for all 1 ≤ m ≤ k for some k ≥ 6. To prove the induction step, we need to prove
         
k+1 (k + 1) + 2 (k + 1) + 4 k+1 (k + 1) + 3
+ + = + .
3 6 6 2 6

Note that
     
k+1 (k + 1) + 2 (k + 1) + 4
+ +
3 6 6
     
k−5 (k − 5) + 2 (k − 5) + 4
= +2 + +1 + +1
3 6 6
     
k−5 (k − 5) + 2 (k − 5) + 4
= +2+ +1+ +1
3 6 6
     
k−5 (k − 5) + 2 (k − 5) + 4
= + + + 4.
3 6 6

By assumption (30.1) when m = k − 5,


         
k−5 (k − 5) + 2 (k − 5) + 4 k−5 (k − 5) + 3
+ + = + .
3 6 6 2 6

68
Therefore,
     
k+1 (k + 1) + 2 (k + 1) + 4
+ +
3 6 6
   
k−5 (k − 5) + 3
= + +4
2 6
   
k−5 (k − 5) + 3
= +3+ +1
2 6
   
k−5 (k − 5) + 3
= +3 + +1
2 6
   
k+1 (k + 1) + 3
= + .
2 6

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1, n = 2, n = 3, n = 4, n = 5, and n = 6. Then,
we proved that it is true for n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k.
Therefore, by strong induction,
jnk n + 2 n + 4 jnk n + 3
+ + = +
3 6 6 2 6

for all n ∈ N.

69
Problem 31
Topic: Sequence

Problem: Let a1 , a2 be arbitrary real numbers, and

an = 2an−1 − an−2

for n ≥ 3. Prove that

an = (n − 1)a2 − (n − 2)a1

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. We can check that

(1 − 1)a2 − (1 − 2)a1 = a1

and
(2 − 1)a2 − (2 − 2)a1 = a2 ,
so it is true for n = 1 and n = 2.

Induction Step
Assume
am = (m − 1)a2 − (m − 2)a1 (31.1)
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

ak+1 = ((k + 1) − 1)a2 − ((k + 1) − 2)a1 = ka2 − (k − 1)a1 .

71
Since k + 1 ≥ 3, we can use the recursive formula and get

ak+1 = 2ak − ak−1 .

By assumption (31.1) when m = k,

ak = (k − 1)a2 − (k − 2)a1 .

By assumption (31.1) when m = k − 1,

ak−1 = (k − 2)a2 − (k − 3)a1 .

So,

ak+1 = 2[(k − 1)a2 − (k − 2)a1 ] − [(k − 2)a2 − (k − 3)a1 ]


= ka2 − (k − 1)a1 .

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

an = (n − 1)a2 − (n − 2)a1

for all n ∈ N.

72
Problem 32
Topic: Sequence

Problem: Let a1 = 5, a2 = 11, a3 = 29, and

an = 3an−1 + an−2 − 3an−3

for n ≥ 4. Prove that


an = 3n + 2
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1, n = 2, and n = 3. We can check that

a1 = 5 = 31 + 2,

a2 = 11 = 32 + 2,
and
a3 = 29 = 33 + 2,
so it is true for n = 1, n = 2, and n = 3.

Induction Step
Assume
am = 3m + 2 (32.1)
for all 1 ≤ m ≤ k for some k ≥ 3. To prove the induction step, we need to prove

ak+1 = 3k+1 + 2.

73
Since k + 1 ≥ 4, we can use the recursive formula and get

ak+1 = 3ak + ak−1 − 3ak−2 .

By assumption (32.1) when m = k,

ak = 3k + 2.

By assumption (32.1) when m = k − 1,

ak−1 = 3k−1 + 2.

By assumption (32.1) when m = k − 2,

ak−2 = 3k−2 + 2.

So,

ak+1 = 3 3k + 2 + 3k−1 + 2 − 3 3k−2 + 2


  

= 3k+1 + 2.

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1, n = 2, and n = 3. Then, we proved that it is
true for n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong
induction,
an = 3n + 2
for all n ∈ N.

74
Problem 33
Topic: Sequence

Problem: Let a1 , a2 , a3 be arbitrary real numbers, and

an = 3an−1 − 3an−2 + an−3

for n ≥ 4. Prove that


1 1
an = (n − 1)(n − 2)a3 − (n − 1)(n − 3)a2 + (n − 2)(n − 3)a1
2 2
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1, n = 2, and n = 3. We can check that
1 1
(1 − 1)(1 − 2)a3 − (1 − 1)(1 − 3)a2 + (1 − 2)(1 − 3)a1 = a1 ,
2 2
1 1
(2 − 1)(2 − 2)a3 − (2 − 1)(2 − 3)a2 + (2 − 2)(2 − 3)a1 = a2 ,
2 2
and
1 1
(3 − 1)(3 − 2)a3 − (3 − 1)(3 − 3)a2 + (3 − 2)(3 − 3)a1 = a3 ,
2 2
so it is true for n = 1, n = 2, and n = 3.

Induction Step
Assume
1 1
am = (m − 1)(m − 2)a3 − (m − 1)(m − 3)a2 + (m − 2)(m − 3)a1 (33.1)
2 2

75
for all 1 ≤ m ≤ k for some k ≥ 3. To prove the induction step, we need to prove
1 1
ak+1 = k(k − 1)a3 − k(k − 2)a2 + (k − 1)(k − 2)a1 .
2 2
Since k + 1 ≥ 4, we can use the recursive formula and get

ak+1 = 3ak − 3ak−1 + ak−2 .

By assumption (33.1) when m = k,


1 1
ak = (k − 1)(k − 2)a3 − (k − 1)(k − 3)a2 + (k − 2)(k − 3)a1 .
2 2
By assumption (33.1) when m = k − 1,
1 1
ak−1 = (k − 2)(k − 3)a3 − (k − 2)(k − 4)a2 + (k − 3)(k − 4)a1 .
2 2
By assumption (33.1) when m = k − 2,
1 1
ak−2 = (k − 3)(k − 4)a3 − (k − 3)(k − 5)a2 + (k − 4)(k − 5)a1 .
2 2
So,
3 3
ak+1 = (k − 1)(k − 2)a3 − 3(k − 1)(k − 3)a2 + (k − 2)(k − 3)a1
2 2
3 3
− (k − 2)(k − 3)a3 + 3(k − 2)(k − 4)a2 − (k − 3)(k − 4)a1
2 2
1 1
+ (k − 3)(k − 4)a3 − (k − 3)(k − 5)a2 + (k − 4)(k − 5)a1
2 2
1 1
= k(k − 1)a3 − k(k − 2)a2 + (k − 1)(k − 2)a1 .
2 2
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1, n = 2, and n = 3. Then, we proved that it is
true for n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong
induction,
1 1
an = (n − 1)(n − 2)a3 − (n − 1)(n − 3)a2 + (n − 2)(n − 3)a1
2 2
for all n ∈ N.

76
Problem 34
Topic: Sequence, Inequality

Problem: Let a0 = 1 and an = nan−1 + n for n ≥ 1. Prove that

an ≥ 3 · n!

for n ≥ 2.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 2. Using the recursive formula,

a1 = 1a0 + 1 = 1 · 1 + 1 = 2,

and
a2 = 2a1 + 2 = 2 · 2 + 2 = 6 ≥ 3 · 2!,
so it is true for n = 2.

Induction Step
Assume
ak ≥ 3 · k! (34.1)

77
for some k ≥ 2. To prove the induction step, we need to prove

ak+1 ≥ 3 · (k + 1)!.

Using the recursive formula and by assumption (34.1),

ak+1 = (k + 1)ak + (k + 1)
≥ (k + 1) · 3 · k! + k + 1
≥ (k + 1) · 3 · k!
= 3 · (k + 1)!.

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 2. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

an ≥ 3 · n!

for n ≥ 2.

78
Problem 35
Topic: Sequence, Inequality

Problem: Let a0 = 1 and an = nan−1 + r for n ≥ 1 where r is some non-


negative real number. Prove that

an ≥ (1 + r) · n!

for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

a1 = 1a0 + r = 1 · 1 + r = 1 + r ≥ (1 + r) · 1!,

so it is true for n = 1.

Induction Step
Assume
ak ≥ (1 + r) · k! (35.1)

79
for some k ≥ 1. To prove the induction step, we need to prove

ak+1 ≥ (1 + r) · (k + 1)!.

Using the recursive formula and by assumption (35.1), since r is non-negative,

ak+1 = (k + 1)ak + r
≥ (k + 1) · (1 + r) · k! + r
≥ (k + 1) · (1 + r) · k!
= (1 + r) · (k + 1)!.

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

an ≥ (1 + r) · n!

for all n ∈ N.

80
Problem 36
Topic: Sequence, Inequality

Problem: Let a0 be an arbitrary real number such that 0 < a0 < 1. Let
r
1 + an−1
an =
2
for n ≥ 1. Prove that
1
1− < an < 1
2n
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. Using the recursive formula,
r
1 + a0
a1 = .
2
Since 0 < a0 < 1,

1 < 1 + a0 < 2
1 1 + a0
< <1
2 r 2
1 1 + a0
√ < < 1.
2 2
√ √
Since 2< 4 = 2,
1 1
√ > .
2 2

81
So,
1 1
1− 1
= < a1 < 1,
2 2
and it is true for n = 1.

Induction Step
Assume
1
1−
< ak < 1 (36.1)
2k
for some k ≥ 1. To prove the induction step, we need to prove
1
1 − k+1 < ak+1 < 1.
2
Using the recursive formula,
r
1 + ak
ak+1 = .
2
By assumption (36.1),
1
2− < 1 + ak < 2
2k
1 1 + ak
1 − k+1 < <1
r 2 r2
1 1 + ak
1 − k+1 < < 1,
2 2
so r
1
1 − k+1 < ak+1 < 1.
2
1
Since 1 − 2k+1
< 1,
r
1
1− < 1,
2k+1
q
1
and multiplying 1− 2k+1
to both sides gives
r
1 1
1 − k+1 < 1 − k+1 .
2 2
So,
1
1−
< ak+1 < 1.
2k+1
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
1
1 − n < an < 1
2
for all n ∈ N.

82
Problem 37
Topic: Sequence

Problem: Let a0 be an arbitrary real number, and

an = 2a2n−1

for n ≥ 1. Prove that n


−1 2n
an = 22 a0
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. Using the recursive formula,

a1 = 2a20 .

We can check that


1
−1 21
22 a0 = 2a20 = a1 ,
so it is true for n = 1.

Induction Step
Assume
k
−1 2k
ak = 22 a0 (37.1)
for some k ≥ 1. To prove the induction step, we need to prove
k+1
−1 2k+1
ak+1 = 22 a0 .

83
Using the recursive formula and the assumption (37.1),

ak+1 = 2a2k
k 2
 k 
= 2 22 −1 a20
k
−2 2·2k
= 2 · 22·2 a0
k+1
−1 2k+1
= 22 a0 .

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
−1 2n
an = 22 a0

for all n ∈ N.

84
Problem 38
Topic: Sequence

Problem: Let a0 = −2 and

(n + 2)an−1
an =
n + 1 + 2an−1

for n ≥ 1. Prove that


n+2
an =
2n − 1
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. By the recursive formula of an ,

(1 + 2)a0 3 · (−2) 1+2


a1 = = =3= ,
1 + 1 + 2a0 2 + 2 · (−2) 2·1−1

so it is true for n = 1.

Induction Step
Assume
k+2
ak = (38.1)
2k − 1
for some k ≥ 1. To prove the induction step, we need to prove

(k + 1) + 2 k+3
ak+1 = = .
2(k + 1) − 1 2k + 1

85
Using the recursive formula of an and the assumption (38.1),

(k + 3)ak
ak+1 =
k + 2 + 2ak
k+2
(k + 3) · 2k−1
= k+2
k+2+2· 2k−1
(k + 3)(k + 2)
=
(k + 2)(2k − 1) + 2(k + 2)
(k + 3)(k + 2)
=
(k + 2)(2k + 1)
k+3
= .
2k + 1
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n+2
an =
2n − 1
for all n ∈ N.

86
Problem 39
Topic: Sequence, Floor Function, Inequality

Problem: Let a0 be an arbitrary real number, and

ban−1 c an−1
an = +
2 2
for n ≥ 1 where b·c denotes the floor function. Prove that
n
a0 − < an ≤ a0
2
for all n ∈ N.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. By the recursive formula,
ba0 c a0
a1 = + .
2 2

87
From the definition of floor function, we know that

ba0 c ≤ a0 < ba0 c + 1 ⇔ a0 − 1 < ba0 c ≤ a0 ,

so
a0 − 1 ba0 c a0
< ≤
2 2 2
1 ba0 c a0
a0 − < + ≤ a0 .
2 2 2
Hence,
1
a0 − < a1 ≤ a0 ,
2
and it is true for n = 1.

Induction Step
Assume
k
a0 − < ak ≤ a0 (39.1)
2
for some k ≥ 1. To prove the induction step, we need to prove
k+1
a0 − < ak+1 ≤ a0 .
2
By the recursive formula,
bak c ak
ak+1 = + .
2 2
From the definition of floor function, we know that

bak c ≤ ak < bak c + 1 ⇔ ak − 1 < bak c ≤ ak ,

so
ak − 1 bak c ak
< ≤
2 2 2
1 bak c ak
ak − < + ≤ ak ,
2 2 2
and we have
1
ak − < ak+1 ≤ ak .
2
Using assumption (39.1),
ak+1 ≤ ak ≤ a0 ,
and
1 k 1 k+1
ak+1 > ak − > a0 − − = a0 − .
2 2 2 2
Hence,
k+1
a0 − < ak+1 ≤ a0 .
2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

88
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
a0 − < an ≤ a0
2
for all n ∈ N.

89
Problem 40
Topic: Sequence, Floor Function, Inequality

Problem: Let a1 , b1 be arbitrary real numbers such that a1 > b1 . Let

ban−1 c 1
an = +
3 2
and
bn−1 1
bn = +
3 6
for n ≥ 2. Prove that an > bn for all n ∈ N.

Definition - Floor Function: For x ∈ R, the floor function bxc is the largest
integer not exceeding x. For example, b2c = 2, b3.6c = 3, b−1.2c = −2, etc.
Formally, for any ` ∈ Z,

bxc = ` ⇔ ` ≤ x < ` + 1.

With this definition of floor function, it should be easy to show that bxc has
the property
bx + ac = bxc + a
for any a ∈ Z.

Solution
Base Cases
We need to prove that it is true for n = 1. Since we are given that a1 > b1 , it is
true for n = 1.

90
Induction Step
Assume
ak > bk (40.1)
for some k ≥ 1. To prove the induction step, we need to prove ak+1 > bk+1 . By
the recursive formula of an ,
bak c 1
ak+1 = + .
3 2
From the definition of floor function, we know that

ak < bak c + 1 ⇔ bak c > ak − 1.

So,
ak − 1 1 ak 1
ak+1 > + = + .
3 2 3 6
Using assumption (40.1), since ak > bk ,

bk 1
ak+1 > + .
3 6
By the recursive formula of bn ,
bk 1
+ = bk+1 .
3 6
Hence, ak+1 > bk+1 . Thus, we proved that it is true for n = k + 1 if it is true for
n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

an > bn

for all n ∈ N.

91
Problem 41
Topic: Harmonic Number, Inequality

Problem: Prove
Hn Hn(2) + 2 ≥ Hn + Hn(2) + Hn(3)
(m)
for all n ∈ N where Hn denotes the n-th Harmonic number and Hn denotes
the n-th generalized Harmonic number of order m.

Definition - Harmonic Number:


n
X 1
Hn = .
j=1
j

Definition - Generalized Harmonic Number:


n
X 1
Hn(m) = m
.
j=1
j

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
(2) (2) (3)
H1 H1 + 2 = 1 · 1 + 2 = 3 ≥ 3 = 1 + 1 + 1 = H1 + H1 + H1 ,

so it is true for n = 1.

93
Induction Step
Assume
(2) (2) (3)
Hk Hk + 2 ≥ Hk + Hk + Hk (41.1)

for some k ≥ 1. To prove the induction step, we need to prove


(2) (2) (3)
Hk+1 Hk+1 + 2 ≥ Hk+1 + Hk+1 + Hk+1 .

Note that
k+1 k
X 1 X1 1 1
Hk+1 = = + = Hk + ,
j=1
j j=1
j k+1 k+1

k+1 k
(2)
X 1 X 1 1 (2) 1
Hk+1 = 2
= 2
+ 2
= Hk + ,
j=1
j j=1
j (k + 1) (k + 1)2

and
k+1 k
(3)
X 1 X 1 1 (3) 1
Hk+1 = 3
= 3
+ 3
= Hk + .
j=1
j j=1
j (k + 1) (k + 1)3

Using assumption (41.1),


  
(2) 1 (2) 1
Hk+1 Hk+1 +2= Hk + Hk + +2
k+1 (k + 1)2
(2)
(2) Hk Hk 1
= Hk Hk + + + +2
k + 1 (k + 1)2 (k + 1)3
(2)
(2) (3) Hk Hk 1
≥ Hk + Hk + Hk + + + +
k + 1 (k + 1)2 (k + 1)3
(2)
Hk (2) Hk (3)
= Hk + + Hk + + Hk+1 .
k+1 (k + 1)2

For k ≥ 1,
1 1
Hk = 1 + + · · · + ≥ 1,
2 k
and
(2) 1 1
Hk =1+ + · · · + 2 ≥ 1.
22 k
So,

(2) 1 (2) 1 (3)


Hk+1 Hk+1 + 2 ≥ Hk + + Hk + + Hk+1
k+1 (k + 1)2
(2) (3)
= Hk+1 + Hk+1 + Hk+1 .

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

94
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

Hn Hn(2) + 2 ≥ Hn + Hn(2) + Hn(3)

for all n ∈ N.

95
Problem 42
Topic: Harmonic Number, Inequality

Problem: Prove
j
n X
X i+j
Hn Hn(2) <
j=1 i=1
i2 j 2
(m)
for all n ∈ N where Hn denotes the n-th Harmonic number and Hn denotes
the n-th generalized Harmonic number of order m.

Definition - Harmonic Number:


n
X 1
Hn = .
j=1
j

Definition - Generalized Harmonic Number:


n
X 1
Hn(m) = m
.
j=1
j

Solution
Base Cases
We need to prove that it is true for n = 1. We have
(2)
H1 H1 =1·1=1

96
and
j
1 X 1
X i+j X i+1 1+1
= = = 2.
j=1 i=1
i2 j 2 i=1
i2 12 12 12
So,
j
1 X
(2)
X i+j
H1 H1 < ,
j=1 i=1
i2 j 2
and it is true for n = 1.

Induction Step
Assume
j
k X
(2)
X i+j
Hk Hk < (42.1)
j=1 i=1
i2 j 2
for some k ≥ 1. To prove the induction step, we need to prove
k+1 j
(2)
XX i+j
Hk+1 Hk+1 < .
j=1 i=1
i2 j 2

Note that
k+1 j k j k+1
XX i+j X X i + j X i + (k + 1)
= +
j=1 i=1
i2 j 2 j=1 i=1
i2 j 2 i=1
i2 (k + 1)2
j
k X k+1 k+1
X i+j 1 X1 1 X 1
= + +
j=1 i=1
i2 j 2 (k + 1)2 i=1 i k + 1 i=1 i2
j
k X (2)
X i+j Hk+1 H
= + + k+1 .
j=1 i=1
i2 j 2 (k + 1) 2 k+1

Using assumption (42.1),


k+1 j (2)
XX i+j (2) Hk+1 H
2 2
> Hk Hk + 2
+ k+1
j=1 i=1
i j (k + 1) k+1
   
(2) 1 1 1 (2) 1
= Hk Hk + Hk + + Hk +
(k + 1)2 k+1 k+1 (k + 1)2
(2)
(2) Hk H 2
= Hk Hk + 2
+ k +
(k + 1) k + 1 (k + 1)3
(2)
(2) Hk H 1
> Hk Hk + + k +
(k + 1)2 k + 1 (k + 1)3
  
1 (2) 1
= Hk + Hk +
k+1 (k + 1)2
(2)
= Hk+1 Hk+1 .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

97
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
j
n X
X i+j
Hn Hn(2) <
j=1 i=1
i2 j 2

for all n ∈ N.

98
Problem 43
Topic: Harmonic Number, Inequality

Problem: Prove
n
H2n ≥ 1 +
2
for all n ∈ N where Hn denotes the n-th Harmonic number.

Definition - Harmonic Number:


n
X 1
Hn = .
j=1
j

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1 1
H21 = 1 + ≥1+ ,
2 2
so it is true for n = 1.

Induction Step
Assume
k
H2k ≥ 1 + (43.1)
2
for some k ≥ 1. To prove the induction step, we need to prove
k+1
H2k+1 ≥ 1 + .
2

99
Note that
k+1 k k
2X 2X+2
1 1
H2k+1 = =
j=1
j j=1
j
1 1 1 1 1 1
= + + ··· + k + k + + ··· + k
1 2 2 2 + 1 2k + 2 2 + 2k
1 1 1
= H2k + k + + ··· + k .
2 + 1 2k + 2 2 + 2k
By assumption (43.1),

k 1 1 1
H2k+1 ≥ 1 + + k + k + ··· + k .
2 2 +1 2 +2 2 + 2k

Since 2k + 1 ≤ 2k + 2 ≤ · · · ≤ 2k + 2k ,
1 1 1
≥ k ≥ ··· ≥ k .
2k + 1 2 +2 2 + 2k
So,
k 1 1 1
H2k+1 ≥ 1 + + + k + ··· + k
2 2k + 2k 2 + 2k 2 + 2k
| {z }
2k terms
k
k 2
=1+ + k
2 2 + 2k
k 1
=1+ +
2 2
k+1
=1+ .
2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
H2n ≥ 1 +
2
for all n ∈ N.

100
Problem 44
Topic: Series, Trigonometry

Problem: Prove
n n+1
 n

X cos 2 sin 2
cos j = 1
 .
j=1
sin 2

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1 1+1
 1

X cos 2 sin 2
cos j = cos 1 = 1
 ,
j=1
sin 2

so it is true for n = 1.

Induction Step
Assume
k k+1
 k

X cos 2 sin 2
cos j = 1
 (44.1)
j=1
sin 2

for some k ≥ 1. To prove the induction step, we need to prove

k+1 k+2
 k+1

X cos 2 sin 2
cos j = 1
 .
j=1
sin 2

101
Using assumption (44.1),
k+1
X k
X
cos j = cos j + cos (k + 1)
j=1 j=1
k+1 k
 
cos 2 sin 2
= 1
 + cos (k + 1).
sin 2
1

Multiplying both sides by sin 2 ,
  k+1      
1 X k+1 k 1
sin cos j = cos sin + cos (k + 1) sin .
2 j=1 2 2 2

Using the formula


1
cos (a) sin (b) = [sin (a + b) − sin (a − b)],
2
we have         
k+1 k 1 1 1
cos sin = sin k + − sin
2 2 2 2 2
and       
1 1 3 1
cos (k + 1) sin = sin k + − sin k + .
2 2 2 2
Hence,
          
k+1 k 1 1 3 1
cos sin + cos (k + 1) sin = sin k + − sin
2 2 2 2 2 2
   
k+2 k+1
= cos sin ,
2 2
and we have
  k+1    
1 X k+2 k+1
sin cos j = cos sin
2 j=1 2 2
k+1
cos k+2 sin k+1
 
X
2  2 .
cos j = 1
j=1
sin 2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
cos n+1 sin n2
 
X
2
cos j =
sin 12

j=1

for all n ∈ N.

102
Problem 45
Topic: Series, Trigonometry

Problem: Prove
n n+1
 n

X sin 2 sin 2
sin j = 1
 .
j=1
sin 2

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1 1+1
 1

X sin 2 sin 2
sin j = sin 1 = 1
 ,
j=1
sin 2

so it is true for n = 1.

Induction Step
Assume
k k+1
 k

X sin 2 sin 2
sin j = 1
 (45.1)
j=1
sin 2

for some k ≥ 1. To prove the induction step, we need to prove

k+1 k+2
 k+1

X sin 2 sin 2
sin j = 1
 .
j=1
sin 2

103
Using assumption (45.1),
k+1
X k
X
sin j = sin j + sin (k + 1)
j=1 j=1
k+1 k
 
sin 2 sin 2
= 1
 + sin (k + 1).
sin 2
1

Multiplying both sides by sin 2 ,
  k+1      
1 X k+1 k 1
sin sin j = sin sin + sin (k + 1) sin .
2 j=1 2 2 2

Using the formula


1
sin (a) sin (b) = [cos (a − b) − cos (a + b)],
2
we have         
k+1 k 1 1 1
sin sin = cos − cos k +
2 2 2 2 2
and       
1 1 1 3
sin (k + 1) sin = cos k + − cos k + .
2 2 2 2
Hence,
          
k+1 k 1 1 1 3
sin sin + sin (k + 1) sin = cos − cos k +
2 2 2 2 2 2
   
k+2 k+1
= sin sin ,
2 2
and we have
  k+1    
1 X k+2 k+1
sin sin j = sin sin
2 j=1 2 2
k+1
sin k+2 sin k+1
 
X
2  2 .
sin j = 1
j=1
sin 2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
sin n+1 sin n2
 
X
2
sin j =
sin 12

j=1

for all n ∈ N.

104
Problem 46
Topic: Fibonacci Number, Golden Number

Problem: Prove
ϕn = Fn ϕ + Fn−1
for all n ∈ N where ϕ is the golden number and Fn denotes the n-th Fibonacci
number.

Definition - Golden Number: The golden number ϕ is the positive real


solution of x2 − x − 1 = 0. Its exact value is

1+ 5
ϕ= .
2

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

F1 ϕ + F0 = ϕ = ϕ1

because F1 = 1 and F0 = 0, so it is true for n = 1.

105
Induction Step
Assume
ϕk = Fk ϕ + Fk−1 (46.1)
for some k ≥ 1. To prove the induction step, we need to prove

ϕk+1 = Fk+1 ϕ + Fk .

Multiplying both sides of (46.1) by ϕ,

ϕk+1 = Fk ϕ2 + Fk−1 ϕ.

From the definition of ϕ, we have

ϕ2 − ϕ − 1 = 0 ⇔ ϕ2 = ϕ + 1.

So,

ϕk+1 = Fk (ϕ + 1) + Fk−1 ϕ
= (Fk−1 + Fk )ϕ + Fk .

Using the recursive relation of Fn ,

Fk−1 + Fk = Fk+1 .

Hence,
ϕk+1 = Fk+1 ϕ + Fk .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

ϕn = Fn ϕ + Fn−1

for all n ∈ N.

106
Problem 47
Topic: Fibonacci Number, Golden Number

Problem: Prove Binet’s formula for Fn ,

ϕn − (1 − ϕ)n
Fn = √ ,
5
where Fn denotes the n-th Fibonacci number and ϕ is the golden number.

Definition - Golden Number: The golden number ϕ is the positive real


solution of x2 − x − 1 = 0. Its exact value is

1+ 5
ϕ= .
2

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,
 √ 
1+ 5
1
ϕ − (1 − ϕ) 1
2ϕ − 1 2· 2 −1
√ = √ = √ = 1 = F1 .
5 5 5

107
When n = 2,

ϕ2 − (1 − ϕ)2 ϕ2 − 1 − 2ϕ + ϕ2 2ϕ − 1
√ = √ = √ = 1 = 1 + 0 = F1 + F0 = F2 .
5 5 5
So, it is true for n = 1 and n = 2.

Induction Step
Assume
ϕm − (1 − ϕ)m
Fm = √ (47.1)
5
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

ϕk+1 − (1 − ϕ)k+1
Fk+1 = √ .
5
Using the recursive relation of Fn ,

Fk+1 = Fk−1 + Fk .

By assumption (47.1) when m = k − 1,

ϕk−1 − (1 − ϕ)k−1
Fk−1 = √ .
5
By assumption (47.1) when m = k,

ϕk − (1 − ϕ)k
Fk = √ .
5
So,

ϕk−1 − (1 − ϕ)k−1 ϕk − (1 − ϕ)k


Fk+1 = √ + √
5 5
ϕk−1 (1 + ϕ) − (1 − ϕ)k−1 (1 + 1 − ϕ)
= √ .
5
From the definition of ϕ, we have

ϕ2 − ϕ − 1 = 0 ⇔ ϕ2 = ϕ + 1.

We can check that 1 − ϕ is also a solution of x2 − x − 1 = 0 since

(1 − ϕ)2 − (1 − ϕ) − 1 = 1 − 2ϕ + ϕ2 − 1 + ϕ − 1
= ϕ2 − ϕ − 1 = 0.

So,
(1 − ϕ)2 = 1 − ϕ + 1.

108
Hence,

ϕk−1 · ϕ2 − (1 − ϕ)k−1 · (1 − ϕ)2


Fk+1 = √
5
ϕk+1 − (1 − ϕ)k+1
= √ .
5
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

ϕn − (1 − ϕ)n
Fn = √
5
for all n ∈ N.

109
Problem 48
Topic: Lucas Number, Golden Number

Problem: Prove Binet’s formula for Ln ,

Ln = ϕn + (1 − ϕ)n ,

where Ln denotes the n-th Lucas number and ϕ is the golden number.

Definition - Golden Number: The golden number ϕ is the positive real


solution of x2 − x − 1 = 0. Its exact value is

1+ 5
ϕ= .
2

Definition - Lucas Numbers: The Lucas numbers are a sequence of numbers


defined as L0 = 2, L1 = 1, and Ln = Ln−1 + Ln−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,

ϕ1 + (1 − ϕ)1 = 1 = L1 .

When n = 2,

ϕ2 + (1 − ϕ)2 = ϕ2 + 1 − 2ϕ + ϕ2 = 2ϕ2 − 2ϕ + 1.


110
From the definition of ϕ, we have

ϕ2 − ϕ − 1 = 0 ⇔ ϕ2 = ϕ + 1,

so
ϕ2 + (1 − ϕ)2 = 2(ϕ + 1) − 2ϕ + 1 = 3 = 2 + 1 = L0 + L1 = L2 .
So, it is true for n = 1 and n = 2.

Induction Step
Assume
Lm = ϕm + (1 − ϕ)m (48.1)
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

Lk+1 = ϕk+1 + (1 − ϕ)k+1 .

Using the recursive relation of Ln ,

Lk+1 = Lk−1 + Lk .

By assumption (48.1) when m = k − 1,

Lk−1 = ϕk−1 + (1 − ϕ)k−1 .

By assumption (48.1) when m = k,

Lk = ϕk + (1 − ϕ)k .

So,

Lk+1 = ϕk−1 + (1 − ϕ)k−1 + ϕk + (1 − ϕ)k


= ϕk−1 (1 + ϕ) + (1 − ϕ)k−1 (1 + 1 − ϕ).

From the definition of ϕ, we have

ϕ2 − ϕ − 1 = 0 ⇔ ϕ2 = ϕ + 1.

We can check that 1 − ϕ is also a solution of x2 − x − 1 = 0 since

(1 − ϕ)2 − (1 − ϕ) − 1 = 1 − 2ϕ + ϕ2 − 1 + ϕ − 1
= ϕ2 − ϕ − 1 = 0.

So,
(1 − ϕ)2 = 1 − ϕ + 1.
Hence,

Lk+1 = ϕk−1 · ϕ2 + (1 − ϕ)k−1 · (1 − ϕ)2


= ϕk+1 + (1 − ϕ)k+1 .

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

111
Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

Ln = ϕn + (1 − ϕ)n

for all n ∈ N.

112
Problem 49
Topic: Fibonacci Number, Lucas Number

Problem: Prove
Ln = Fn−1 + Fn+1
where Ln denotes the n-th Lucas number and Fn denotes the n-th Fibonacci
number.

Definition - Lucas Numbers: The Lucas numbers are a sequence of numbers


defined as L0 = 2, L1 = 1, and Ln = Ln−1 + Ln−2 for n ≥ 2.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,
F0 + F2 = F0 + F0 + F1 = 0 + 0 + 1 = 1 = L1 .
When n = 2,
F1 + F3 = F1 + F1 + F2 = F1 + F1 + F0 + F1 = 1 + 1 + 0 + 1 = 3
and
L2 = L0 + L1 = 2 + 1 = 3,
so
F1 + F3 = L2 .
So, it is true for n = 1 and n = 2.

113
Induction Step
Assume
Lm = Fm−1 + Fm+1 (49.1)
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

Lk+1 = Fk + Fk+2 .

Using the recursive relation of Ln ,

Lk+1 = Lk−1 + Lk .

By assumption (49.1) when m = k − 1,

Lk−1 = Fk−2 + Fk .

By assumption (49.1) when m = k,

Lk = Fk−1 + Fk+1 .

So,

Lk+1 = Fk−2 + Fk + Fk−1 + Fk+1


= (Fk−2 + Fk−1 ) + (Fk + Fk+1 ).

Using the recursive relation of Fn ,

Fk−2 + Fk−1 = Fk ,

and
Fk + Fk+1 = Fk+2 .
Hence,
Lk+1 = Fk + Fk+2 .
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

Ln = Fn−1 + Fn+1

for all n ∈ N.

114
Problem 50
Topic: Sequence, Fibonacci Number

Problem: Let a1 and a2 be arbitrary real numbers, and

an = an−1 + an−2

for n ≥ 3. Prove that


an = a1 Fn−2 + a2 Fn−1
for n ≥ 2 where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 2 and n = 3. When n = 2,

a2 = a1 F0 + a2 F1

because F0 = 0 and F1 = 1. When n = 3, by the recursive formula of an ,

a3 = a1 + a2 = a1 F1 + a2 F2

because F1 = 1 and F2 = F0 + F1 = 0 + 1 = 1. So, it is true for n = 2 and n = 3.

Induction Step
Assume
am = a1 Fm−2 + a2 Fm−1 (50.1)

115
for all 2 ≤ m ≤ k for some k ≥ 3. To prove the induction step, we need to prove

ak+1 = a1 Fk−1 + a2 Fk .

By the recursive formula of an ,

ak+1 = ak−1 + ak .

By assumption (50.1) when m = k − 1,

ak−1 = a1 Fk−3 + a2 Fk−2 .

By assumption (50.1) when m = k,

ak = a1 Fk−2 + a2 Fk−1 .

So,

ak+1 = a1 Fk−3 + a2 Fk−2 + a1 Fk−2 + a2 Fk−1


= a1 (Fk−3 + Fk−2 ) + a2 (Fk−2 + Fk−1 ).

Using the recursive relation of Fn ,

Fk−3 + Fk−2 = Fk−1 ,

and
Fk−2 + Fk−1 = Fk .
Hence,
ak+1 = a1 Fk−1 + a2 Fk .
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 2 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 2 and n = 3. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 2 ≤ m ≤ k. Therefore, by strong induction,

an = a1 Fn−2 + a2 Fn−1

for n ≥ 2.

116
Problem 51
Topic: Fibonacci Number

Problem: For `, n ∈ N, prove

F`+n = F`−1 Fn + F` Fn+1

where Fn denotes the n-th Fibonacci number.

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1, by the recursive
relation of Fn ,
F`+1 = F`−1 + F` = F`−1 F1 + F` F2
because F1 = 1 and F2 = F0 + F1 = 0 + 1 = 1. When n = 2, by the recursive
relation of Fn ,

F`+2 = F` + F`+1 = F` + F`−1 + F` = F`−1 + 2F` = F`−1 F2 + F` F3

because F2 = 1 and F3 = F1 + F2 = 1 + 1 = 2. So, it is true for n = 1 and n = 2.

Induction Step
Assume
F`+m = F`−1 Fm + F` Fm+1 (51.1)

118
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

F`+k+1 = F`−1 Fk+1 + F` Fk+2 .

Using the recursive relation of Fn ,

F`+k+1 = F`+k−1 + F`+k .

By assumption (51.1) when m = k − 1,

F`+k−1 = F`−1 Fk−1 + F` Fk .

By assumption (51.1) when m = k,

F`+k = F`−1 Fk + F` Fk+1 .

So,

F`+k+1 = F`−1 Fk−1 + F` Fk + F`−1 Fk + F` Fk+1


= F`−1 (Fk−1 + Fk ) + F` (Fk + Fk+1 ).

Using the recursive relation of Fn ,

Fk−1 + Fk = Fk+1 ,

and
Fk + Fk+1 = Fk+2 .
Hence,
F`+k+1 = F`−1 Fk+1 + F` Fk+2 .
Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

F`+n = F`−1 Fn + F` Fn+1

for all n ∈ N.

119
Problem 52
Topic: Iterated Function

Problem: For a, b ∈ R and a 6= 0, if f (x) = ax + b, then prove


n−1
X
f n (x) = an x + b aj
j=0

for all n ∈ N where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1−1
X
f 1 (x) = a1 x + b aj = ax + b = f (x),
j=0

so it is true for n = 1.

120
Induction Step
Assume
k−1
X
f k (x) = ak x + b aj (52.1)
j=0

for some k ≥ 1. To prove the induction step, we need to prove


k
X
f k+1 (x) = ak+1 x + b aj .
j=0

Using assumption (52.1),


k−1
X
f k+1 (x) = f k ◦ f (x) = ak (ax + b) + b aj

j=0
k−1
X
= ak+1 x + bak + b aj
j=0
k
X
= ak+1 x + b aj .
j=0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n−1
X
f n (x) = an x + b aj
j=0

for all n ∈ N.

121
Problem 53
Topic: Iterated Function, Fibonacci Number

Problem: If
1
f (x) = ,
1+x
then prove
Fn−1 x + Fn
f n (x) =
Fn x + Fn+1
for all n ∈ N where f n (x) denotes the n-th iterate of f (x) and Fn denotes the
n-th Fibonacci number.

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
F0 x + F1 0·x+1 1
f 1 (x) = = = = f (x)
F1 x + F2 1·x+1 1+x

122
because F0 = 0, F1 = 1, and F2 = F0 + F1 = 0 + 1 = 1, so it is true for n = 1.

Induction Step
Assume
Fk−1 x + Fk
f k (x) = (53.1)
Fk x + Fk+1
for some k ≥ 1. To prove the induction step, we need to prove
Fk x + Fk+1
f k+1 (x) = .
Fk+1 x + Fk+2

Using assumption (53.1),


1
Fk−1 · 1+x + Fk
f k+1 (x) = f k ◦ f (x) =

1
Fk · 1+x + Fk+1
1
Fk−1 · 1+x + Fk 1+x
= 1 ·
Fk · 1+x + Fk+1 1+x
Fk−1 + Fk + Fk x
= .
Fk + Fk+1 + Fk+1 x

Using the recursive relation of Fn ,

Fk−1 + Fk = Fk+1 ,

and
Fk + Fk+1 = Fk+2 .
Hence,
Fk x + Fk+1
f k+1 (x) = .
Fk+1 x + Fk+2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
Fn−1 x + Fn
f n (x) =
Fn x + Fn+1

for all n ∈ N.

123
Problem 54
Topic: Iterated Function

Problem: If p
f (x) = x2 + 1,
then prove p
f n (x) = x2 + n
for all n ∈ N where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
p
f 1 (x) = x2 + 1 = f (x),

so it is true for n = 1.

Induction Step
Assume p
f k (x) = x2 + k (54.1)

124
for some k ≥ 1. To prove the induction step, we need to prove
p
f k+1 (x) = x2 + k + 1.

Using assumption (54.1),


r 2
p
k+1 k

f (x) = f ◦ f (x) = x2 + 1 + k
p
= x2 + k + 1.

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
p
f n (x) = x2 + n

for all n ∈ N.

125
Problem 55
Topic: Iterated Function

Problem: For a ∈ R and some function g, if

f (x) = g −1 (g(x) + a),

then prove
f n (x) = g −1 (g(x) + na)
for all n ∈ N where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
f 1 (x) = g −1 (g(x) + 1 · a) = g −1 (g(x) + a) = f (x),
so it is true for n = 1.

Induction Step
Assume
f k (x) = g −1 (g(x) + ka) (55.1)

126
for some k ≥ 1. To prove the induction step, we need to prove

f k+1 (x) = g −1 (g(x) + (k + 1)a).

Using assumption (55.1),

f k+1 (x) = f k ◦ f (x)




= g −1 (g(f (x)) + ka).

Since
g(f (x)) = g(g −1 (g(x) + a)) = g(x) + a,
we have

f k+1 (x) = g −1 (g(x) + a + ka)


= g −1 (g(x) + (k + 1)a).

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

f n (x) = g −1 (g(x) + na)

for all n ∈ N.

127
Problem 56
Topic: Iterated Function

Problem: If
f (x) = 2x(1 − x),
then prove
1h n
i
f n (x) = 1 − (1 − 2x)2
2
n
for all n ∈ N where f (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1h 1
i
f 1 (x) = 1 − (1 − 2x)2
2
1
4x − 4x2

=
2
= 2x − 2x2
= 2x(1 − x) = f (x),
so it is true for n = 1.

128
Induction Step
Assume
1h k
i
f k (x) =
1 − (1 − 2x)2 (56.1)
2
for some k ≥ 1. To prove the induction step, we need to prove
1h k+1
i
f k+1 (x) = 1 − (1 − 2x)2 .
2
Using assumption (56.1),
1h k
i
f k+1 (x) = f k ◦ f (x) = 1 − (1 − 2 · 2x(1 − x))2

2 
1 2k
= 1 − 1 − 4x + 4x2
2
1h k+1
i
= 1 − (1 − 2x)2 .
2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
1h n
i
f n (x) = 1 − (1 − 2x)2
2
for all n ∈ N.

129
Problem 57
Topic: Iterated Function

Problem: If
f (x) = 2x2 − 1,
then prove

1  p 2n  p 2n 
f n (x) = x + x2 − 1 + x − x2 − 1
2

for all n ∈ N where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1  p 21  p 21 
1 2 2
f (x) = x+ x −1 + x− x −1
2
1h 2 p p i
= x + 2x x2 − 1 + (x2 − 1) + x2 − 2x x2 − 1 + (x2 − 1)
2
= 2x2 − 1 = f (x),
so it is true for n = 1.

130
Induction Step
Assume 
1  p 2k  p 2k 
k 2 2
f (x) = x+ x −1 + x− x −1 (57.1)
2
for some k ≥ 1. To prove the induction step, we need to prove

1  p 2k+1  p 2k+1 
k+1 2 2
f (x) = x+ x −1 + x− x −1 .
2

Using assumption (57.1),

f k+1 (x)
= f k ◦ f (x)

" 2k  2k #
1
q q
2 2
= 2x2 − 1 + (2x2 − 1) − 1 + 2x2 − 1 − (2x2 − 1) − 1
2

1  2 p 2k  p 2k 
2 2 2
= 2x − 1 + 2x x − 1 + 2x − 1 − 2x x − 1 .
2

Note that p  p 2
2x2 − 1 + 2x x2 − 1 = x + x2 − 1

and p  p 2
2x2 − 1 − 2x x2 − 1 = x − x2 − 1 .

Hence, 
1  p 2k+1  p 2k+1 
k+1 2 2
f (x) = x+ x −1 + x− x −1 .
2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

1  p 2n  p 2n 
n 2 2
f (x) = x+ x −1 + x− x −1
2

for all n ∈ N.

131
Problem 58
Topic: Iterated Function, Derivative

Problem: If f (x) = ex , then prove


n
d n Y
f (x) = f j (x)
dx j=1

for all n ∈ N where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1
d 1 d d x Y
f (x) = f (x) = e = ex = f 1 (x) = f j (x),
dx dx dx j=1

so it is true for n = 1.

132
Induction Step
Assume
k
d k Y
f (x) = f j (x) (58.1)
dx j=1

for some k ≥ 1. To prove the induction step, we need to prove


k+1
d k+1 Y
f (x) = f j (x).
dx j=1

By applying the chain rule of derivative and using assumption (58.1),

d k+1 d
f k ◦ f (x)

f (x) =
dx dx
d k x
= f (e )
dx
k
Y d x
= f j (ex ) × e
j=1
dx
k
Y
= f j (ex ) × ex .
j=1

Since f j (ex ) = f j ◦ f (x) = f j+1 (x) and ex = f 1 (x),

k
d k+1 Y
f (x) = f j+1 (x) × f 1 (x)
dx j=1
k+1
Y
= f j (x) × f 1 (x)
j=2
k+1
Y
= f j (x).
j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
d n Y
f (x) = f j (x)
dx j=1

for all n ∈ N.

133
Problem 59
Topic: Iterated Function, Derivative

Problem: If f (x) = ln(x), then prove


 −1
n−1
d n 1 Y j 
f (x) = f (x)
dx x j=1

for n ≥ 2 where f n (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 2. Applying the chain rule of derivative,
 −1
2−1
d 2 d 1 1 1 1 1 Y j 
f (x) = ln(ln(x)) = × = 1 × = f (x) ,
dx dx ln(x) x f (x) x x j=1

so it is true for n = 2.

134
Induction Step
Assume  −1
k−1
d k 1 Y
f (x) =  f j (x) (59.1)
dx x j=1

for some k ≥ 2. To prove the induction step, we need to prove


 −1
k
d k+1 1 Y
f (x) =  f j (x) .
dx x j=1

By applying the chain rule of derivative and using assumption (59.1),


d k+1 d
f k ◦ f (x)

f (x) =
dx dx
d k
= f (ln(x))
dx
 −1
k−1
1  Y d
= f j (ln(x)) × ln(x)
ln(x) j=1 dx
 −1
k−1
1 Y j 1
= f (ln(x)) × .
ln(x) j=1 x

Since f j (ln(x)) = f j ◦ f (x) = f j+1 (x) and ln(x) = f 1 (x),
 −1
k−1
d k+1 1  Y 1
f (x) = 1 f j+1 (x) ×
dx f (x) j=1 x
 −1
k
1 Y j  1
= 1 f (x) ×
f (x) j=2 x
 −1
k
1 Y
= f j (x) .
x j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 2. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 −1
n−1
d n 1 Y
f (x) =  f j (x)
dx x j=1

for n ≥ 2.

135
Problem 60
Topic: Iterated Function

Problem: If
x2
f (x) = ,
2x − 1
then prove
 −1
n−1
Yh i
n j j
f n (x) = x2  x2 + (x − 1)2 
j=0

n
for all n ∈ N where f (x) denotes the n-th iterate of f (x).

Definition - Iterated Function: The n-th iterate of f (x) is the composition


of f (x) with itself n times, i.e.

f n (x) = (f ◦ f ◦ · · · ◦ f )(x).
| {z }
n times

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
 −1
1−1 h i
1 Y j j
f 1 (x) = x2  x2 + (x − 1)2 
j=0
h 0 0
i−1 x2
= x2 x2 + (x − 1)2 = = f (x),
2x − 1

136
so it is true for n = 1.

Induction Step
Assume  −1
k−1
Yh i
k j j
f k (x) = x2  x2 + (x − 1)2  (60.1)
j=0

for some k ≥ 1. To prove the induction step, we need to prove


 −1
k h i
k+1 Y j j
f k+1 (x) = x2  x2 + (x − 1)2  .
j=0

Using assumption (60.1),

f k+1 (x)
= f k ◦ f (x)


2j # −1
 
2k k−1
" 2j
x2 x2 x2
 Y 
=  + −1 
2x − 1 j=0
2x − 1 2x − 1

2j # −1
 
 2k −1  2
2k k−1
"
2
2j  2
2x − 1 x Y x x
=  + −1 
2x − 1 2x − 1 j=0
2x − 1 2x − 1

2j # −1
" j
 
k+1 k−1
Y  x2 2
x2 2

1 x
= ·  + −1  .
(2x − 1)2k −1 2x − 1 j=0 2x − 1 2x − 1

In problem 11, we proved that


n−1
X an − 1
aj = .
j=0
a−1

Let a = 2 and n = k,
k−1
X
k
2 −1= 2j = 20 + 21 + · · · + 2k−2 + 2k−1 .
j=0

So,
1 1 1 1 1
k −1 = 0 · 1 ··· k−2 ·
(2x − 1)2 (2x − 1) 2 (2x − 1) 2 (2x − 1) 2 (2x − 1)2k−1
 −1
k−1
Y j
=  (2x − 1)2  .
j=0

137
Hence,

f k+1 (x)
2j # −1 k−1
 " j
  −1
k+1 k−1
Y  x2 2
x2 2

x Y j
  (2x − 1)2 
=  + −1
2x − 1 j=0 2x − 1 2x − 1 j=0
 −1
2k+1 k−1
Y 
x j
2 2
2j
=  x + x2 − 2x + 1 
2x − 1 j=0
 −1
k+1 k−1
x2 Y h j+1 2j+1
i
=  x2 + (x − 1) 
2x − 1 j=0
 −1
2k+1 k h
x Y j 2 j
i
=  x2 + (x − 1)  .
2x − 1 j=1

Since
1 0 −1
h 0 i
= x2 + (x − 1)2 ,
2x − 1
we have  −1
k h i
k+1 Y j j
f k+1 (x) = x2  x2 + (x − 1)2  .
j=0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 −1
n−1
Yh i
n j j
f n (x) = x2  x2 + (x − 1)2 
j=0

for all n ∈ N.

138
Problem 61
Topic: Product

Problem: Prove
n  j
Y 1 (n + 1)n
1+ = .
j=1
j n!

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1  j 1
(1 + 1)1

Y 1 1
1+ = 1+ =2= ,
j=1
j 1 1!

so it is true for n = 1.

Induction Step
Assume
k  j
Y 1 (k + 1)k
1+ = (61.1)
j=1
j k!

140
for some k ≥ 1. To prove the induction step, we need to prove
k+1 j
Y 1 (k + 2)k+1
1+ = .
j=1
j (k + 1)!

Using assumption (61.1),


k+1
Y j  k+1 Y k  j
1 1 1
1+ = 1+ × 1+
j=1
j k+1 j=1
j
k+1
(k + 1)k

k+2
= ×
k+1 k!
(k + 2)k+1
=
(k + 1) · k!
(k + 2)k+1
= .
(k + 1)!

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n  j
Y 1 (n + 1)n
1+ =
j=1
j n!

for all n ∈ N.

141
Problem 62
Topic: Product, Trigonometry

Problem: Prove
n x
Y sin x
cos = .
j=1
2j 2n sin 2xn

Solution
Base Cases
We need to prove that it is true for n = 1. Using the identity sin(2θ) = 2 sin (θ) cos(θ),
x x
sin(x) = 2 sin cos ,
2 2
so
1 x x
Y sin x
cos j = cos = x .

j=1
2 2 2 sin 2

So, it is true for n = 1.

Induction Step
Assume
k x
Y sin x
cos = (62.1)
sin 2xk

j=1
2j 2k

for some k ≥ 1. To prove the induction step, we need to prove


k+1 x
Y sin x
cos = x
.
j=1
2j 2k+1 sin 2k+1

142
Using assumption (62.1),
k+1
Y x  x  Y k x
cos = cos × cos
j=1
2j 2k+1 j=1
2j
 x  sin x
= cos k+1 × k .
2 2 sin 2xk

Using the identity sin(2θ) = 2 sin (θ) cos(θ),


x

x  x   x   x  sin 2k 
sin k
= 2 sin k+1 cos k+1 ⇔ cos k+1 = x ,
2 2 2 2 2 sin 2k+1

so
k+1 x

Y x sin 2k  sin x
cos = x ×
2k sin 2xk

j=1
2j 2 sin 2k+1
sin x
= x
.
2k+1 sin 2k+1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n x
Y sin x
cos =
2n sin 2xn

j=1
2j

for all n ∈ N.

143
Problem 63
Topic: Product, Series

Problem: Prove n
n−1
Y  2X −1
2j
1+x = xj .
j=0 j=0

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
1−1
Y  2X −1
j 0
1 + x2 = 1 + x2 = 1 + x = x0 + x1 = xj ,
j=0 j=0

so it is true for n = 1.

Induction Step
Assume
k
k−1
Y  2X −1
j
1 + x2 = xj (63.1)
j=0 j=0

for some k ≥ 1. To prove the induction step, we need to prove

k   2k+1
X−1
2j
Y
1+x = xj .
j=0 j=0

144
Using assumption (63.1),
k   k−1
Y   
Y j j k
1 + x2 = 1 + x2 × 1 + x2
j=0 j=0
 k

2X −1  
k
= xj  × 1 + x2
j=0
k k
2X −1 2X −1
k
= xj + xj+2 .
j=0 j=0

Changing the index j + 2k → j,


k
2X −1 2k −1+2k
2k+1
X−1
j+2k
X
j
x = x = xj .
j=0 j=2k j=2k

Hence,
k
k   2X −1 2k+1
X−1 2k+1
X−1
2j
Y
j j
1+x = x + x = xj .
j=0 j=0 j=2k j=0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
n−1
Y  2X −1
2j
1+x = xj
j=0 j=0

for all n ∈ N.

145
Problem 64
Topic: Product, Series

Problem: Prove n
n−1
Y  3X −1
3j 2·3j
1+x +x = xj .
j=0 j=0

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
1−1
Y  3X −1
j j 0 0
1 + x3 + x2·3 = 1 + x3 + x2·3 = 1 + x + x2 = x0 + x1 + x2 = xj ,
j=0 j=0

so it is true for n = 1.

Induction Step
Assume
k
k−1
Y  3X −1
j j
1 + x3 + x2·3 = xj (64.1)
j=0 j=0

for some k ≥ 1. To prove the induction step, we need to prove

k   3k+1
X−1
3j 2·3j
Y
1+x +x = xj .
j=0 j=0

146
Using assumption (64.1),
k   k−1
Y   
Y j j j j k k
1 + x3 + x2·3 = 1 + x3 + x2·3 × 1 + x3 + x2·3
j=0 j=0
 k

3X −1  
k k
= xj  × 1 + x3 + x2·3
j=0
k k k
3X −1 3X −1 3X −1
k k
= xj + xj+3 + xj+2·3 .
j=0 j=0 j=0

Changing the index j + 3k → j,


k
3X −1 3k −1+3k k
X−1
2·3
j+3k
X
j
x = x = xj .
j=0 j=3k j=3k

Changing the index j + 2 · 3k → j,


k
3X −1 3k −1+2·3k
3k+1
X−1
j+2·3k
X
j
x = x = xj .
j=0 j=2·3k j=2·3k

Hence,
k k
k   3X −1 X−1
2·3 3k+1
X−1 3k+1
X−1
3j 2·3j
Y
j j j
1+x +x = x + x + x = xj .
j=0 j=0 j=3k j=2·3k j=0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
n−1
Y  3X −1
3j 2·3j
1+x +x = xj
j=0 j=0

for all n ∈ N.

147
Problem 65
Topic: Supergolden Number, Narayana Sequence

Problem: Prove that

ψ n = Nn−2 ψ 2 + Nn−4 ψ + Nn−3

for n ≥ 4 where ψ is the supergolden number and Nn denotes the n-th number
in Narayana sequence.

Definition - Supergolden Number: The supergolden number ψ is the real


solution of x3 − x2 − 1 = 0. Its exact value is
√ √
 s s 
1 3 29 + 3 93 3 29 − 3 93 
ψ= 1+ + .
3 2 2

Definition - Narayana Sequence: The Narayana sequence {Nn } is a se-


quence of numbers defined by N0 = N1 = N2 = 1 and Nn = Nn−1 + Nn−3 for
n ≥ 3.

Solution
Base Cases
We need to prove that it is true for n = 4. From the definition of supergolden
number,
ψ 3 − ψ 2 − 1 = 0 ⇔ ψ 3 = ψ 2 + 1.

148
Multiplying both sides by ψ,

ψ 4 = ψ 3 + ψ = ψ 2 + ψ + 1 = N2 ψ 2 + N0 ψ + N1

because N0 = N1 = N2 = 1. So, it is true for n = 4.

Induction Step
Assume
ψ k = Nk−2 ψ 2 + Nk−4 ψ + Nk−3 (65.1)
for some k ≥ 4. To prove the induction step, we need to prove

ψ k+1 = Nk−1 ψ 2 + Nk−3 ψ + Nk−2 .

Multiplying both sides of (65.1) by ψ,

ψ k+1 = Nk−2 ψ 3 + Nk−4 ψ 2 + Nk−3 ψ.

From the definition of supergolden number,

ψ 3 − ψ 2 − 1 = 0 ⇔ ψ 3 = ψ 2 + 1.

So,

ψ k+1 = Nk−2 ψ 2 + 1 + Nk−4 ψ 2 + Nk−3 ψ




= (Nk−2 + Nk−4 ) ψ 2 + Nk−3 ψ + Nk−2 .

Using the recursive relation of Nn ,

Nk−2 + Nk−4 = Nk−1 .

Hence,
ψ k+1 = Nk−1 ψ 2 + Nk−3 ψ + Nk−2 .
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 4. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

ψ n = Nn−2 ψ 2 + Nn−4 ψ + Nn−3

for n ≥ 4.

149
Problem 66
Topic: Inequality

Problem: Prove
n
X n
X
xj ≤ |xj |
j=1 j=1

for all n ∈ N for real numbers xj , which is an extension of the triangle inequal-
ity.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,

1
X 1
X
xj = |x1 | ≤ |x1 | = |xj |,
j=1 j=1

so it is true for n = 1. Now using the fact that a number is less than or equal to
its absolute value,

x1 x2 ≤ |x1 x2 |
x1 x2 ≤ |x1 ||x2 |.

Multiplying both sides by 2,

2x1 x2 ≤ 2|x1 ||x2 |.

Adding x21 + x22 to both sides,

x21 + 2x1 x2 + x22 ≤ x21 + 2|x1 ||x2 | + x22 .

150
Since x21 = |x1 |2 and x22 = |x2 |2 ,

x21 + 2x1 x2 + x22 ≤ |x1 |2 + 2|x1 ||x2 | + |x2 |2


2
(x1 + x2 )2 ≤ (|x1 | + |x2 |) .

Taking square root of both sides,

|x1 + x2 | ≤ |x1 | + |x2 |.

So, it is true for n = 2.

Induction Step
Assume
m
X m
X
xj ≤ |xj | (66.1)
j=1 j=1

for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

k+1
X k+1
X
xj ≤ |xj |.
j=1 j=1

By assumption (66.1) when m = 2,

|x1 + x2 | ≤ |x1 | + |x2 |.


k
X
Let x1 = xj and x2 = xk+1 ,
j=1

k
X k
X
xj + xk+1 ≤ xj + |xk+1 |
j=1 j=1

k+1
X k
X
xj ≤ xj + |xk+1 |.
j=1 j=1

By assumption (66.1) when m = k,

k
X k
X
xj ≤ |xj |.
j=1 j=1

Hence,
k+1
X k
X k+1
X
xj ≤ |xj | + |xk+1 | = |xj |.
j=1 j=1 j=1

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

151
Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

n
X n
X
xj ≤ |xj |
j=1 j=1

for all n ∈ N.

152
Problem 67
Topic: Inequality

Problem: Prove Cauchy-Schwarz inequality,

(a1 b1 + a2 b2 + · · · + an bn )2 ≤ a21 + a22 + · · · + a2n b21 + b22 + · · · + b2n ,


 

for all n ∈ N and a1 , · · · , an , b1 , · · · , bn ≥ 0.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,

(a1 b1 )2 = a21 b21 ≤ a21 b21 ,


 

so it is true for n = 1. When n = 2,

a21 + a22 b21 + b22 − (a1 b1 + a2 b2 )2 = a21 b22 + a22 b21 − 2a1 b1 a2 b2
 

= (a1 b2 − a2 b1 )2 ≥ 0,

so it is true that
(a1 b1 + a2 b2 )2 ≤ a21 + a22 b21 + b22 .
 

Induction Step
Assume

(a1 b1 + a2 b2 + · · · + am bm )2 ≤ a21 + a22 + · · · + a2m b21 + b22 + · · · + b2m


 
q q
⇔a1 b1 + a2 b2 + · · · + am bm ≤ a21 + a22 + · · · + a2m b21 + b22 + · · · + b2m (67.1)

153
for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove

(a1 b1 + a2 b2 + · · · + ak+1 bk+1 )2 ≤ a21 + a22 + · · · + a2k+1 b21 + b22 + · · · + b2k+1 .


 

By assumption (67.1) when m = k,

a1 b1 + a2 b2 + · · · + ak bk + ak+1 bk+1
q q
≤ a21 + a22 + · · · + a2k b21 + b22 + · · · + b2k + ak+1 bk+1 . (67.2)

By assumption (67.1) when m = 2,


q q
a1 b1 + a2 b2 ≤ a21 + a22 b21 + b22 .
p p
Let a1 = a21 + a22 + · · · + a2k , b1 = b21 + b22 + · · · + b2k , a2 = ak+1 , and b2 = bk+1 ,
q q
a21 + a22 + · · · + a2k b21 + b22 + · · · + b2k + ak+1 bk+1
q q
≤ a21 + a22 + · · · + a2k + a2k+1 b21 + b22 + · · · + b2k + b2k+1 . (67.3)

Hence, from (67.2) and (67.3),


q q
a1 b1 + a2 b2 + · · · + ak+1 bk+1 ≤ a21 + a22 + · · · + a2k+1 b21 + b22 + · · · + b2k+1

(a1 b1 + a2 b2 + · · · + ak+1 bk+1 )2 ≤ a21 + a22 + · · · + a2k+1 b21 + b22 + · · · + b2k+1 .


 

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,

(a1 b1 + a2 b2 + · · · + an bn )2 ≤ a21 + a22 + · · · + a2n b21 + b22 + · · · + b2n


 

for all n ∈ N.

154
Problem 68
Topic: Inequality

Problem: Prove Aczél’s inequality, which states that if

x21 + x22 + · · · + x2n < x20

and
y12 + y22 + · · · + yn2 < y02 ,
then

x20 − x21 − x22 − · · · − x2n y02 − y12 − y22 − · · · − yn2


 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xn yn ) .

Solution
Base Cases
We need to prove that it is true for n = 1, i.e. if x21 < x20 and y12 < y02 , then
2
x20 − x21 y02 − y12 ≤ (x0 y0 − x1 y1 ) .
 

Since
2
(x0 y0 − x1 y1 ) − x20 − x21 y02 − y12 = −2x0 y0 x1 y1 + x21 y02 + x20 y12
 

= (x1 y0 − x0 y1 )2 ≥ 0,

it is true that
2
x20 − x21 y02 − y12 ≤ (x0 y0 − x1 y1 ) .
 

155
Induction Step
Assume if x21 + x22 + · · · + x2m < x20 and y12 + y22 + · · · + ym
2
< y02 , then

x20 − x21 − x22 − · · · − x2m y02 − y12 − y22 − · · · − ym


2
 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xm ym ) (68.1)

for all 1 ≤ m ≤ k for some k ≥ 1. To prove the induction step, we need to prove
that if x21 + x22 + · · · + x2k+1 < x20 and y12 + y22 + · · · + yk+1
2
< y02 , then

x20 − x21 − x22 − · · · − x2k+1 y02 − y12 − y22 − · · · − yk+1


2
 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xk+1 yk+1 ) .

First, note that


q 2
x21 + x22 + · · · + x2k+1 < x20 ⇒ x21 + x22 + · · · + x2k < x20 − x2k+1 (68.2)

and
q 2
y12 + y22 + · · · + yk+1
2
< y02 ⇒ y12 + y22 + · · · + yk2 < y02 − yk+1
2 . (68.3)

By assumption (68.1) when m = k, if x21 +x22 +· · ·+x2k < x20 and y12 +y22 +· · ·+yk2 <
y02 , then

x20 − x21 − x22 − · · · − x2k y02 − y12 − y22 − · · · − yk2


 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xk yk ) .
q q
Let x0 = x20 − x2k+1 and y0 = y02 − yk+1 2 , if

q 2
x21 + x22 + · · · + x2k < x20 − x2k+1

and q 2
y12 + y22 + · · · + yk2 < y02 − yk+1
2 ,

then

x20 − x2k+1 − x21 − x22 − · · · − x2k y02 − yk+1


2
− y12 − y22 − · · · − yk2
 
q q 2
≤ x20 − x2k+1 y02 − yk+1
2 − x1 y1 − x2 y2 − · · · − xk yk . (68.4)

By assumption (68.1) when m = 1, if x21 < x20 and y12 < y02 , then
2
x20 − x21 y02 − y12 ≤ (x0 y0 − x1 y1 ) .
 

Let x1 = xk+1 and y1 = yk+1 , if x2k+1 < x20 and yk+1


2
< y02 , then
2
x20 − x2k+1 y02 − yk+1
2
 
≤ (x0 y0 − xk+1 yk+1 ) ,

156
which implies q q
x20 − x2k+1 y02 − yk+1
2 ≤ x0 y0 − xk+1 yk+1 .
So,
q q 2
x20 − x2k+1 y02 − yk+1
2 − x1 y1 − x2 y2 − · · · − xk yk
2
≤ (x0 y0 − xk+1 yk+1 − x1 y1 − x2 y2 − · · · − xk yk ) . (68.5)

Hence, from (68.2), (68.3), (68.4), and (68.5), if x21 + x22 + · · · + x2k+1 < x20 and
y12 + y22 + · · · + yk+1
2
< y02 , then

x20 − x21 − x22 − · · · − x2k+1 y02 − y12 − y22 − · · · − yk+1


2
 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xk+1 yk+1 ) .

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction, if
x21 + x22 + · · · + x2n < x20 and y12 + y22 + · · · + yn2 < y02 , then

x20 − x21 − x22 − · · · − x2n y02 − y12 − y22 − · · · − yn2


 

2
≤ (x0 y0 − x1 y1 − x2 y2 − · · · − xn yn )

for all n ∈ N.

157
Problem 69
Topic: Inequality

Problem: Prove
n
X n
Y
aj − aj ≤ n − 1
j=1 j=1

if 0 ≤ aj ≤ 1 for all 1 ≤ j ≤ n for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
1
X 1
Y
aj − aj = a1 − a1 = 0 ≤ 1 − 1,
j=1 j=1

so it is true for n = 1.

Induction Step
Assume
k
X k
Y
aj − aj ≤ k − 1 (69.1)
j=1 j=1

if 0 ≤ aj ≤ 1 for all 1 ≤ j ≤ k for some k ≥ 1. To prove the induction step, we


need to prove
k+1
X k+1
Y
aj − aj ≤ (k + 1) − 1 = k
j=1 j=1

158
if 0 ≤ aj ≤ 1 for all 1 ≤ j ≤ k + 1. First, if 0 ≤ aj ≤ 1 for all 1 ≤ j ≤ k + 1, then
a1 + a2 + · · · + ak ≤ k and ak+1 ≤ 1, which imply
 
Xk
k − aj  ≥ 0
j=1

and (1 − ak+1 ) ≥ 0. Multiplying them together,


 
Xk
k − aj  (1 − ak+1 ) ≥ 0
j=1
k
X k
X
k− aj − kak+1 + ak+1 aj ≥ 0
j=1 j=1
k
X k
X
k− aj − ak+1 − (k − 1)ak+1 + ak+1 aj ≥ 0
j=1 j=1
   
Xk Xk
k− aj + ak+1  + ak+1  aj − (k − 1) ≥ 0
j=1 j=1
 
k+1
X Xk
k− aj + ak+1  aj − (k − 1) ≥ 0.
j=1 j=1

So,  
k+1
X Xk
aj ≤ k + ak+1  aj − (k − 1) .
j=1 j=1

By assumption (69.1),
k
X k
Y
aj − (k − 1) ≤ aj ,
j=1 j=1
so
 
k
X k
Y
k + ak+1  aj − (k − 1) ≤ k + ak+1 · aj
j=1 j=1
k+1
Y
=k+ aj .
j=1

Hence,
k+1
X k+1
Y
aj ≤ k + aj
j=1 j=1
k+1
X k+1
Y
aj − aj ≤ k.
j=1 j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

159
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n
X n
Y
aj − aj ≤ n − 1
j=1 j=1

if 0 ≤ aj ≤ 1 for all 1 ≤ j ≤ n for all n ∈ N.

160
Problem 70
Topic: Inequality

Problem: For x1,j , x2,j , · · · , xn,j ≥ 0 for all 1 ≤ j ≤ `, prove


 2     
X` X` X` `
X
 x1,j x2,j · · · xn,j  ≤  x21,j   x22,j  · · ·  x2n,j 
j=1 j=1 j=1 j=1

for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 2. In problem 67, we proved that

 2   
`
X X` `
X
 a j bj  ≤  a2j   b2j  .
j=1 j=1 j=1

Let aj → x1,j and bj → x2,j ,

 2   
X` `
X X`
 x1,j x2,j  ≤  x21,j   x22,j  .
j=1 j=1 j=1

So, it is true for n = 2.

161
Induction Step
Assume
 2     
`
X X` X` X`
 x1,j x2,j · · · xk,j  ≤  x21,j   x22,j  · · ·  x2k,j  (70.1)
j=1 j=1 j=1 j=1

for some k ≥ 2. To prove the induction step, we need to prove


 2     
X ` X` X` `
X
 x1,j x2,j · · · xk+1,j  ≤  x21,j   x22,j  · · ·  x2k+1,j  .
j=1 j=1 j=1 j=1

Using assumption (70.1), let xk,j → xk,j xk+1,j ,


 2     
X` X` X` X`
 x1,j x2,j · · · xk,j xk+1,j  ≤  x21,j   x22,j  · · ·  x2k,j x2k+1,j  .
j=1 j=1 j=1 j=1

By expanding the product of two sums, we have


  
X ` X`
 x2k,j   x2k+1,j 
j=1 j=1
`
X X X
= x2k,j x2k+1,j + x2k,i x2k+1,j + x2k,i x2k+1,j
j=1 1≤i<j≤` 1≤j<i≤`
`
X
≥ x2k,j x2k+1,j .
j=1

Hence,
 2
X`
 x1,j x2,j · · · xk,j xk+1,j 
j=1
     
X` X` `
X X`
≤ x21,j   x22,j  · · ·  x2k,j   x2k+1,j  .
j=1 j=1 j=1 j=1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 2. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 2     
X ` X` X` X`
 x1,j x2,j · · · xn,j  ≤  x21,j   x22,j  · · ·  x2n,j 
j=1 j=1 j=1 j=1

for n ≥ 2.

162
Problem 71
Topic: Inequality

Problem: For aj , bj > 0 for all 1 ≤ j ≤ n, prove Milne’s inequality


     
n n n n
X X a b
j j 
X X
 (aj + bj )  ≤ aj   bj 
j=1 j=1
a j + bj j=1 j=1

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1, we can check
that
     
1 1 1 1
X
 (aj + bj ) 
X a j b j  = a1 b1 ≤ a1 b1 = 
X X
aj   bj  ,
j=1 j=1
a j + b j j=1 j=1

so it is true for n = 1. For n = 2, since


     
2 2 2 2
X X X X aj bj 
 aj   bj  −  (aj + bj ) 
j=1 j=1 j=1 j=1
aj + bj

(a1 + b1 )a2 b2 (a2 + b2 )a1 b1


=a1 b2 + a2 b1 − −
a2 + b2 a1 + b1
2
(a1 b2 − a2 b1 )
= ≥ 0,
(a1 + b1 )(a2 + b2 )

164
it is true that
     
2 2 2 2
X X aj bj  X X
 (aj + bj )  ≤ aj   bj  .
j=1 j=1
aj + bj j=1 j=1

Induction Step
Assume      
m m m m
X
 (aj + bj ) 
X a j bj ≤
X X
aj   bj  (71.1)
j=1 j=1
a j + b j j=1 j=1

for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
     
k+1
X k+1
X aj bj k+1
X k+1
X
 (aj + bj )  ≤ aj   bj  .
j=1
a + bj
j=1 j j=1 j=1

First, note that


  
k+1
X k+1
X a j bj
 (aj + bj )  
j=1 j=1
aj + bj
  
k k
X X a j b j ak+1 bk+1
=  (aj + bj ) + (ak+1 + bk+1 )  + 
j=1 j=1
a j + b j ak+1 + bk+1
  
k k k
X X aj bj  ak+1 bk+1 X
=  (aj + bj )  + ak+1 bk+1 + (aj + bj )
j=1
a + bj
j=1 j
ak+1 + bk+1 j=1
k
X aj bj
+ (ak+1 + bk+1 )
a + bj
j=1 j

and
  
k+1
X k+1
X
 aj   bj 
j=1 j=1
  
Xk k
X
= aj + ak+1   bj + bk+1 
j=1 j=1
  
Xk Xk k
X k
X
=  aj   bj + ak+1 bk+1 + bk+1
 aj + ak+1 bj .
j=1 j=1 j=1 j=1

By assumption (71.1) when m = k,


     
k k k k
X X aj bj  X X
 (aj + bj )  ≤ aj   bj  . (71.2)
j=1 j=1
aj + bj j=1 j=1

165
By assumption (71.1) when m = 2,
 
a1 b1 a2 b2
[(a1 + b1 ) + (a2 + b2 )] + ≤ (a1 + a2 )(b1 + b2 )
a1 + b1 a 2 + b2
(a1 + b1 )a2 b2 (a2 + b2 )a1 b1
a1 b1 + a2 b2 + + ≤ a 1 b1 + a 2 b2 + a 1 b2 + a 2 b1
a2 + b2 a1 + b1
(a1 + b1 )a2 b2 (a2 + b2 )a1 b1
+ ≤ a 1 b2 + a 2 b1 .
a2 + b2 a1 + b1
Let a1 = aj , b1 = bj , a2 = ak+1 , b2 = bk+1 , and then taking the sum from j = 1 to
j = k,
k   X k
X (aj + bj )ak+1 bk+1 (ak+1 + bk+1 )aj bj
+ ≤ (aj bk+1 + ak+1 bj )
j=1
ak+1 + bk+1 aj + bj j=1

k k k k
ak+1 bk+1 X X aj bj X X
(aj + bj ) + (ak+1 + bk+1 ) ≤ bk+1 aj + ak+1 bj .
ak+1 + bk+1 j=1 a + bj
j=1 j j=1 j=1
(71.3)
Adding (71.2) and (71.3) and then adding ak+1 bk+1 to both sides, we have the
inequality
     
k+1
X k+1
X aj bj k+1
X k+1
X
 (aj + bj )  ≤ aj   bj  .
j=1 j=1
aj + b j j=1 j=1

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
     
n n n n
X X a b
j j 
X X
 (aj + bj )  ≤ aj   bj 
j=1 j=1
a j + b j j=1 j=1

for all n ∈ N.

166
Problem 72
Topic: Inequality

Problem: For aj , bj > 0 for all 1 ≤ j ≤ n, prove Milne’s inequality


 2   
n n n 2 2
X X X aj b j
a2j + b2j  

 aj bj  ≤  2 + b2

j=1 j=1 j=1
aj j

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1, we can check
that
 2   
1 1 1 2 2
X X X aj bj
aj bj  = a21 b21 ≤ a21 b21 =  a2j + b2j  

 ,
a2 + b2
j=1 j=1 j=1 j j

so it is true for n = 1. For n = 2, since


    2
2 2 2 2 2
X X a j bj
X
a2j + b2j  

 − aj bj 
a 2 + b2
j=1 j=1 j j j=1
2 2
 2 2 2 2
 2 2
a +b a b a +b a b
= 1 2 1 2 2 2 + 2 2 2 2 1 1 − 2a1 b1 a2 b2
a 2 + b2 a1 + b1
 2   2
a1 + b21 a2 b2 − a22 + b22 a1 b1
= ≥ 0,
(a21 + b21 ) (a22 + b22 )

167
it is true that
 2   
2 2 2 2 2
X X X a b
j j 
a2j + b2j  

 aj bj  ≤  2 + b2 .
j=1 j=1
a
j=1 j j

Induction Step
Assume  2   
m m m 2 2
X X X a b
j j 
a2j + b2j  

 aj bj  ≤  2 + b2 (72.1)
j=1 j=1 j=1
aj j

for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
 2   
k+1 k+1 k+1
X X X a2j b2j
a2j + bj  
2

 aj bj  ≤  .
j=1 j=1 j=1
a2j + b2j

First, note that


 2  2
k+1
X Xk
 aj bj  =  aj bj + ak+1 bk+1 
j=1 j=1
 2
Xk k
X
= aj bj  + a2k+1 b2k+1 + 2ak+1 bk+1 aj bj
j=1 j=1

and
  
k+1
X k+1
X a2j b2j
a2j + b2j  

 
j=1
a2 + b2j
j=1 j
  
k k
X
2 2
 2 2
 X a2j b2j a2k+1 b2k+1
= aj + bj + ak+1 + bk+1   + 2 
j=1 j=1 j
a2 + b2j ak+1 + b2k+1
  
k k 2 2 k
X X aj bj a2 b2 X
a2j + b2j    + a2k+1 b2k+1 + k+1 k+1 a2j + b2j
 
= 2 2 2 2
j=1
a + bj
j=1 j
ak+1 + bk+1 j=1
k
X a2j b2j
+ a2k+1 + b2k+1 .
a2 + b2j
j=1 j

By assumption (72.1) when m = k,


 2   
k k k 2 2
X X X a b
j j 
a2j + b2j  

 aj bj  ≤  2 + b2 . (72.2)
j=1 j=1
a
j=1 j j

168
By assumption (72.1) when m = 2,

 a21 b21 a22 b22


 
(a1 b1 + a2 b2 )2 ≤ a21 + b21 + a22 + b22
 
+
a21 + b21 a22 + b22
 
a2 + b2 a2 b2 a2 + b2 a2 b2
a21 b21 + a22 b22 + 2a1 b1 a2 b2 ≤ a21 b21 + a22 b22 + 1 2 1 2 2 2 + 2 2 2 2 1 1
a 2 + b2 a1 + b1
2 2
 2 2 2 2
 2 2
a +b a b a +b a b
2a1 b1 a2 b2 ≤ 1 2 1 2 2 2 + 2 2 2 2 1 1 .
a 2 + b2 a1 + b1

Let a1 = aj , b1 = bj , a2 = ak+1 , b2 = bk+1 , and then taking the sum from j = 1 to


j = k,
k k
"   #
X X a2j + b2j a2k+1 b2k+1 a2k+1 + b2k+1 a2j b2j
2aj bj ak+1 bk+1 ≤ +
j=1 j=1
a2k+1 + b2k+1 a2j + b2j
k k k
X a2k+1 b2k+1 X 2 2
 2 2
X a2j b2j
2ak+1 bk+1 aj bj ≤ 2 2 a j + b j + ak+1 + b k+1 .
j=1
ak+1 + bk+1 j=1 2
a + b2j
j=1 j
(72.3)

Adding (72.2) and (72.3) and then adding a2k+1 b2k+1 to both sides, we have the
inequality
 2   
k+1
X k+1
X k+1
X a2j b2j
a2j + b2j  

 aj bj  ≤  .
j=1 j=1
a2 + b2j
j=1 j

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
 2   
n n n 2 2
X X X a b
j j 
a2j + b2j  

 aj bj  ≤  2 + b2
j=1 j=1 j=1
aj j

for all n ∈ N.

169
Problem 73
Topic: Binomial Coefficient

Problem: Prove the binomial theorem,


n  
n
X n
(x + y) = xn−j y j ,
j=0
j

for all n ∈ N.

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

1      
X 1 1−j j 1 1 0 1 0 1
x y = x y + x y = x + y = (x + y)1 ,
j=0
j 0 1

so it is true for n = 1.

170
Induction Step
Assume
k  
X k k−j j
(x + y)k = x y (73.1)
j=0
j

for some k ≥ 1. To prove the induction step, we need to prove


k+1
X 
k+1 k + 1 k+1−j j
(x + y) = x y .
j=0
j

Multiplying both sides of (73.1) by x + y,


k   k  
X k X k
(x + y)k+1 = xk+1−j y j + xk−j y j+1
j=0
j j=0
j
k   k−1  
X k k+1−j j X k k−j j+1
= xk+1 + x y + x y + y k+1 .
j=1
j j=0
j

Let j + 1 → j,
k−1
X  k   k  
k k−j j+1 X k X k
x y = xk−(j−1) y j = xk+1−j y j .
j=0
j j=1
j − 1 j=1
j − 1

So,
k    
X k k
(x + y)k+1 = xk+1 + + xk+1−j y j + y k+1 .
j=1
j j−1

Since
   
k k k! k!
+ = +
j j−1 j!(k − j)! (j − 1)!(k − j + 1)!
(k − j + 1) · k! j · k!
= +
j!(k − j + 1)! j!(k − j + 1)!
 
(k + 1)! k+1
= = ,
j!(k + 1 − j)! j
which is known as Pascal’s identity, we have
k  
k+1 k+1
X k+1
(x + y) =x + xk+1−j y j + y k+1 .
j=1
j

Note that  
k + 1 k+1−0 0
xk+1 = x y
0
and  
k+1 k + 1 k+1−(k+1) k+1
y = x y .
k+1

171
Hence,
k+1
X 
k + 1 k+1−j j
(x + y)k+1 = x y .
j=0
j

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n  
X n n−j j
(x + y)n = x y
j=0
j

for all n ∈ N.

172
Problem 74
Topic: Binomial Coefficient

Problem: For n, r ∈ N, prove the Hockey-Stick identity,


n    
X j n+1
= ,
j=r
r r+1

for n ≥ r.

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Solution
Base Cases
We need to prove that it is true for n = r. We can check that

r      
X j r r+1
= =1= ,
j=r
r r r+1

so it is true for n = r.

173
Induction Step
Assume
k    
X j k+1
= (74.1)
j=r
r r+1

for some k ≥ r. To prove the induction step, we need to prove


k+1
X   
j k+2
= .
j=r
r r+1

Using assumption (74.1),


k+1
X  k    
j X j k+1
= +
j=r
r j=r
r r
   
k+1 k+1
= + .
r+1 r

In problem 73, we have calculated that


     
k k k+1
+ = .
j j−1 j

Let k → k + 1 and j → r + 1,
     
k+1 k+1 k+2
+ = .
r+1 r r+1

So,
k+1
X   
j k+2
= .
j=r
r r+1

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = r. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n    
X j n+1
=
j=r
r r+1

for n ≥ r.

174
Problem 75
Topic: Binomial Coefficient, Inequality

Problem: Prove  
2n
≤ 22n−1
n
for all n ∈ N.

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that
 
2 2!
= = 2 ≤ 2 = 22·1−1 ,
1 1! 1!

so it is true for n = 1.

Induction Step
Assume  
2k
≤ 22k−1 (75.1)
k

175
for some k ≥ 1. To prove the induction step, we need to prove
 
2k + 2
≤ 22(k+1)−1 = 22k+1 .
k+1

Note that
 
2k + 2 (2k + 2)!
=
k+1 ((k + 1)!)2
(2k + 2)(2k + 1) (2k)!
= ·
(k + 1)2 (k!)2
 
(2k + 2)(2k + 1) 2k
= .
(k + 1)2 k

Since 2k + 1 ≤ 2k + 2, using assumption (75.1),

(2k + 2)2 2k
   
2k + 2

k+1 (k + 1)2 k
 
2 2k
=2
k
≤ 22 · 22k−1
= 22k+1 .

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 
2n
≤ 22n−1
n

for all n ∈ N.

176
Problem 76
Topic: Binomial Coefficient, Fibonacci Number

Problem: For any non-negative integer `, prove


n  
X n
F`+j = F`+2n
j=0
j

for all n ∈ N where Fn denotes the n-th Fibonacci number.

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1. Since
1      
X 1 1 1
F`+j = F`+0 + F`+1 = F` + F`+1 = F`+2 ,
j=0
j 0 1

it is true for n = 1.

177
Induction Step
Assume
k  
X k
F`+j = F`+2k (76.1)
j=0
j
for some k ≥ 1. To prove the induction step, we need to prove
k+1
X k + 1 
F`+j = F`+2(k+1) = F`+2k+2 .
j=0
j

First, note that


k+1
X k + 1   k    
k+1 X k+1 k+1
F`+j = F`+0 + F`+j + F`+k+1
j=0
j 0 j=1
j k+1
k  
X k+1
= F` + F`+j + F`+k+1 .
j=1
j

In problem 73, we have calculated that


     
k+1 k k
= + .
j j j−1
So,
k+1
X  k    
k+1 X k k
F`+j = F` + + F`+j + F`+k+1
j=0
j j=1
j j−1
k   k  
X k X k
= F` + F`+j + F`+j + F`+k+1 .
j=1
j j=1
j−1

Substituting j − 1 → j,
k   k−1
X k 
X k
F`+j = F`+j+1 .
j=1
j−1 j=0
j

So,
k+1
X  k   k−1
X k 
k+1 X k
F`+j = F` + F`+j + F`+1+j + F`+k+1 .
j=0
j j=1
j j=0
j
Note that
k     k   k  
X k k X k X k
F` + F`+j = F`+0 + F`+j = F`+j
j=1
j 0 j=1
j j=0
j

and
k−1
X  k−1
X k    k  
k k X k
F`+1+j + F`+k+1 = F`+1+j + F`+1+k = F`+1+j .
j=0
j j=0
j k j=0
j

178
Hence,
k+1
X  k   k  
k+1 X k X k
F`+j = F`+j + F`+1+j .
j=0
j j=0
j j=0
j

By assumption (76.1),
k  
X k
F`+j = F`+2k
j=0
j

and
k  
X k
F`+1+j = F`+1+2k ,
j=0
j
so
k+1
X 
k+1
F`+j = F`+2k + F`+1+2k = F`+2k+2
j=0
j

by using the recursive relation of Fn . Thus, we proved that it is true for n = k + 1


if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
n  
X n
F`+j = F`+2n
j=0
j

for all n ∈ N.

179
Problem 77
Topic: Binomial Coefficient, Fibonacci Number

Problem: Prove
n  
X n−j
= Fn+1
j=0
j

for all n ∈ N where Fn denotes the n-th Fibonacci number.

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Definition - Fibonacci Numbers: The Fibonacci numbers are a sequence


of numbers defined as F0 = 0, F1 = 1, and Fn = Fn−1 + Fn−2 for n ≥ 2.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,

1      
X 1−j 1 0
= + = 1 + 0 = 1 = F2
j=0
j 0 1

180
because F2 = F1 + F0 = 1 + 0 = 1. When n = 2,
2        
X 2−j 2 1 0
= + + = 1 + 1 + 0 = 2 = F3
j=0
j 0 1 2

because F3 = F2 + F1 = 1 + 1 = 2. So, it is true for n = 1 and n = 2.

Induction Step
Assume
m  
X m−j
= Fm+1 (77.1)
j=0
j

for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
k+1
X 
k+1−j
= Fk+2 .
j=0
j

Note that
k+1
X    X k    
k+1−j k+1 k+1−j 0
= + +
j=0
j 0 j=1
j k+1
k  
X k+1−j
=1+ .
j=1
j

In problem 73, we have calculated that


     
k+1 k k
= + .
j j j−1

Let k → k − j,      
k+1−j k−j k−j
= + .
j j j−1
So,
k+1
X  k   k  
k+1−j X k−j X k−j
=1+ + .
j=0
j j=1
j j=1
j−1

Note that
k     X k   X k  
X k−j k−0 k−j k−j
1+ = + = .
j=1
j 0 j=1
j j=0
j

Substituting j − 1 → j,
k   k−1
X 
X k−j k−1−j
= .
j=1
j−1 j=0
j

181
Hence,
k+1
X  k   k−1
X 
k+1−j X k−j k−1−j
= + .
j=0
j j=0
j j=0
j

By assumption (77.1) when m = k,


k  
X k−j
= Fk+1 .
j=0
j

By assumption (77.1) when m = k − 1,


k−1
X 
k−1−j
= Fk .
j=0
j

So,
k+1
X 
k+1−j
= Fk+1 + Fk = Fk+2
j=0
j

by using the recursive relation of Fn . Thus, we proved that it is true for n = k + 1


if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
n  
X n−j
= Fn+1
j=0
j

for all n ∈ N.

182
Problem 78
Topic: Derivative

Problem: Prove
dn n
x = n!
dxn
for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

d
x = 1 = 1!,
dx
so it is true for n = 1.

Induction Step
Assume
dk k
x = k! (78.1)
dxk

183
for some k ≥ 1. To prove the induction step, we need to prove

dk+1 k+1
x = (k + 1)!.
dxk+1
Using assumption (78.1),

dk+1 k+1 dk
 
d k+1
k+1
x = k x
dx dx dx
dk
(k + 1)xk

=
dxk
dk
= (k + 1) k xk
dx
= (k + 1) · k!
= (k + 1)!.

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
dn n
x = n!
dxn
for all n ∈ N.

184
Problem 79
Topic: Derivative

Problem: If f (x) is an even function, then prove that f (2n) (x) is even for all
n ∈ N where f (n) (x) denotes the n-th derivative of f (x).

Solution
Base Cases
We need to prove that it is true for n = 1. If f (x) is even, then

f (x) = f (−x).

Differentiating both sides with respect to x twice,

f (1) (x) = −f (1) (−x)


 
f (2) (x) = − −f (2) (−x) = f (2) (−x).

So, f (2) (x) is even.

Induction Step
Assume that f (2k) (x) is even if f (x) is even for some k ≥ 1. To prove the induction
step, we need to prove that f (2k+2) (x) is even. By the assumption that f (2k) (x) is
even,
f (2k) (x) = f (2k) (−x).
Differentiating both sides with respect to x twice,

f (2k+1) (x) = −f (2k+1) (−x)


 
f (2k+2) (x) = − −f (2k+2) (−x) = f (2k+2) (−x).

185
Hence, f (2k+2) (x) is even. Thus, we proved that it is true for n = k + 1 if it is true
for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction, if f (x) is an even function,
then f (2n) (x) is even for all n ∈ N.

186
Problem 80
Topic: Derivative, Trigonometry

Problem: Prove
dn  nπ 
sin x = sin x +
dxn 2
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

d
sin x = cos x
dx
and  π π π
sin x + = sin x cos + cos x sin = cos x,
2 2 2
it is true that
d  π
sin x = sin x + .
dx 2

Induction Step
Assume
dk
 

sin x = sin x + (80.1)
dxk 2
for some k ≥ 1. To prove the induction step, we need to prove

dk+1
 
(k + 1)π
sin x = sin x + .
dxk+1 2

187
Differentiating both sides of (80.1) with respect to x using the chain rule,

dk+1
   
d kπ kπ
sin x = sin x + = cos x + .
dxk+1 dx 2 2

By using the addition formula of sine,


     
kπ kπ π kπ π
cos x + = sin x + cos + cos x + sin
2 2 2 2 2
 
(k + 1)π
= sin x + .
2

Hence,
dk+1
 
(k + 1)π
sin x = sin x + .
dxk+1 2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
dn  nπ 
sin x = sin x +
dxn 2
for all n ∈ N.

188
Problem 81
Topic: Derivative, Trigonometry

Problem: Prove
dn  nπ 
cos x = cos x +
dxn 2
for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

d
cos x = − sin x
dx
and  π π π
cos x + = cos x cos − sin x sin = − sin x,
2 2 2
it is true that
d  π
cos x = cos x + .
dx 2

Induction Step
Assume
dk
 

cos x = cos x + (81.1)
dxk 2
for some k ≥ 1. To prove the induction step, we need to prove

dk+1
 
(k + 1)π
cos x = cos x + .
dxk+1 2

190
Differentiating both sides of (81.1) with respect to x using the chain rule,

dk+1
   
d kπ kπ
cos x = cos x + = − sin x + .
dxk+1 dx 2 2

By using the addition formula of cosine,


     
kπ kπ π kπ π
− sin x + = cos x + cos − sin x + sin
2 2 2 2 2
 
(k + 1)π
= cos x + .
2

Hence,
dk+1
 
(k + 1)π
cos x = cos x + .
dxk+1 2
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
dn  nπ 
cos x = cos x +
dxn 2
for all n ∈ N.

191
Problem 82
Topic: Derivative

Problem: Prove
dn (−1)n−1 (n − 1)!
n
ln x =
dx xn
for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

d 1 (−1)1−1 (1 − 1)!
ln x = = ,
dx x x1
so it is true for n = 1.

Induction Step
Assume
dk (−1)k−1 (k − 1)!
k
ln x = (82.1)
dx xk

192
for some k ≥ 1. To prove the induction step, we need to prove

dk+1 (−1)k k!
ln x = .
dxk+1 xk+1
Differentiating both sides of (82.1) with respect to x,

dk+1 d (−1)k−1 (k − 1)!


k+1
ln x =
dx dx xk
d 1
= (−1)k−1 (k − 1)! ×
dx xk
−k
= (−1)k−1 (k − 1)! × k+1
x
(−1)k k!
= .
xk+1
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

dn (−1)n−1 (n − 1)!
ln x =
dxn xn
for all n ∈ N.

193
Problem 83
Topic: Derivative

Problem: For a, b, c, d ∈ R and b, c 6= 0, prove

dn ax + b n!(−c)n−1 (ad − bc)


n
=
dx cx + d (cx + d)n+1

for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. By the quotient rule,

d ax + b a(cx + d) − c(ax + b) ad − bc 1!(−c)1−1 (ad − bc)


= 2
= 2
= ,
dx cx + d (cx + d) (cx + d) (cx + d)1+1
so it is true for n = 1.

Induction Step
Assume
dk ax + b k!(−c)k−1 (ad − bc)
= (83.1)
dxk cx + d (cx + d)k+1

194
for some k ≥ 1. To prove the induction step, we need to prove

dk+1 ax + b (k + 1)!(−c)k (ad − bc)


= .
dxk+1 cx + d (cx + d)k+2

Differentiating both sides of (83.1) with respect to x,

dk+1 ax + b d k!(−c)k−1 (ad − bc)


k+1
=
dx cx + d dx (cx + d)k+1
d 1
= k!(−c)k−1 (ad − bc) ×
dx (cx + d)k+1
c(−k − 1)
= k!(−c)k−1 (ad − bc) ×
(cx + d)k+2
(k + 1)!(−c)k (ad − bc)
= .
(cx + d)k+2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

dn ax + b n!(−c)n−1 (ad − bc)


n
=
dx cx + d (cx + d)n+1

for all n ∈ N.

195
Problem 84
Topic: Derivative

Problem: Prove
dn 1 (2n)! 1
√ = n ·
dx n
1−x 4 n! (1 − x)n+ 12
for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. We can check that

d 1 1 2! 1
√ = = 1 · ,
dx 1 − x 2(1 − x)3/2 4 · 1! (1 − x)1+ 21

so it is true for n = 1.

Induction Step
Assume
dk 1 (2k)! 1
√ = k · (84.1)
dxk 1 − x 4 k! (1 − x)k+ 21

196
for some k ≥ 1. To prove the induction step, we need to prove

dk+1 1 (2k + 2)! 1


√ = k+1 · .
dx k+1
1−x 4 (k + 1)! (1 − x)k+1+ 21

Differentiating both sides of (84.1) with respect to x,

dk+1 1 d (2k)! 1
√ = ·
dxk+1 1 − x dx 4k k! (1 − x)k+ 12
(2k)! d 1
= k ×
4 k! dx (1 − x)k+ 12
(2k)! k + 12
= k
× 1
4 k! (1 − x)k+ 2 +1
(2k)! (2k + 1)(2k + 2) 1
= k × × 1
4 k! 4(k + 1) (1 − x)k+1+ 2
(2k + 2)! 1
= k+1 · .
4 (k + 1)! (1 − x)k+1+ 12

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

dn 1 (2n)! 1
√ = n ·
dxn 1 − x 4 n! (1 − x)n+ 12

for all n ∈ N.

197
Problem 85
Topic: Derivative, Trigonometry

Problem: Prove
dn −1 (−1)n−1 (n − 1)!
sin n cot−1 x

tan x =
dxn 2
(1 + x )
n/2

for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

sin2 cot−1 x + cos2 cot−1 x = 1


 

and
cos cot−1 x

 = cot cot−1 x = x,

−1
sin cot x
we have
1
sin cot−1 x = √

.
1 + x2

198
So,
d 1
tan−1 x =
dx 1 + x2
1 1
=√ ·√
1 + x2 1 + x2
(−1)1−1 (1 − 1)!
sin cot−1 x .

= 1/2
(1 + x2 )

Induction Step
Assume
dk −1 (−1)k−1 (k − 1)!
sin k cot−1 x

tan x = (85.1)
dxk 2
(1 + x )
k/2

for some k ≥ 1. To prove the induction step, we need to prove


dk+1 (−1)k k!
tan−1 x = sin (k + 1) cot−1 x .

dx k+1 (k+1)/2
(1 + x2 )
Differentiating both sides of (85.1) with respect to x using the product rule,

dk+1 −1 d (−1)k−1 (k − 1)!


sin k cot−1 x

k+1
tan x = k/2
dx dx (1 + x2 )
d 1
=(−1)k−1 (k − 1)! sin k cot−1 x ×

dx (1 + x2 )k/2
(−1)k−1 (k − 1)! d
sin k cot−1 x .

+ k/2
×
(1 + x2 ) dx

Using the chain rule,


d 1 −kx
k/2
= (k+2)/2
dx (1 + x2 ) (1 + x2 )
and
d −k
sin k cot−1 x = cos k cot−1 x ·
 
.
dx 1 + x2
So,
dk+1 −kx
tan−1 x =(−1)k−1 (k − 1)! sin k cot−1 x ×

dx k+1 (k+2)/2
(1 + x2 )
(−1)k−1 (k − 1)! −k cos k cot−1 x

+ ×
(1 + x2 )
k/2 1 + x2
(−1)k k!

x
sin k cot−1 x

= (k+1)/2

2 1+x 2
(1 + x )

1
cos k cot−1 x .

+√
1 + x2

199
Since
sin2 cot−1 x + cos2 cot−1 x = 1
 

and
cos cot−1 x

= cot cot−1 x = x,

−1

sin cot x
we have
1
sin cot−1 x = √

1 + x2
and
x
cos cot−1 x = √

.
1 + x2
Hence,

dk+1 −1 (−1)k k!
cos cot−1 x sin k cot−1 x
  
k+1
tan x = (k+1)/2
dx (1 + x2 )
+ sin cot−1 x cos k cot−1 x
 

(−1)k k!
sin (k + 1) cot−1 x .

= (k+1)/2
2
(1 + x )

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

dn −1 (−1)n−1 (n − 1)!
sin n cot−1 x

tan x =
dxn 2
(1 + x )
n/2

for all n ∈ N.

200
Problem 86
Topic: Derivative, Hyperbolic Functions

Problem: Prove
d2n (2n − 1)!
tanh−1 x = −1

2n 2 n sinh 2n tanh x
dx (1 − x )

for all n ∈ N.

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. The second derivative of tanh−1 x is
d2 d 1 2x
tanh−1 x = = 2.
dx2 dx 1 − x2 (1 − x2 )
The hyperbolic cosine cosh x and the hyperbolic sine sinh x satisfy the identity

cosh2 x − sinh2 x = 1,

which can be proven using the definitions


ex + e−x
cosh x =
2

201
and
ex − e−x
sinh x = .
2
Since
cosh2 tanh−1 x − sinh2 tanh−1 x = 1
 

and
sinh tanh−1 x

 = tanh tanh−1 x = x,

−1
cosh tanh x
we have
x
sinh tanh−1 x = √

1 − x2
and
1
cosh tanh−1 x = √

.
1 − x2
Using the identity sinh (2x) = 2 sinh x cosh x,

sinh 2 tanh−1 x = 2 sinh tanh−1 x cosh tanh−1 x


  

x 1 2x
=2· √ ·√ = .
1−x 2 1−x 2 1 − x2

So,
(2 · 1 − 1)! 1 2x 2x
sinh 2 tanh−1 x =

1 2
· 2
= 2.
(1 − x2 ) 1−x 1−x (1 − x2 )
Hence, it is true that

d2 (2 · 1 − 1)!
tanh−1 x = −1

dx2 2 1 sinh 2 tanh x .
(1 − x )

Induction Step
Assume
d2k (2k − 1)!
tanh−1 x = sinh 2k tanh−1 x

2k k
(86.1)
dx (1 − x2 )
for some k ≥ 1. To prove the induction step, we need to prove

d2k+2 (2k + 1)!


tanh−1 x = sinh (2k + 2) tanh−1 x .

dx 2k+2 2 k+1
(1 − x )

Differentiating both sides of (86.1) with respect to x using the product rule,

d2k+1 d 1
tanh−1 x =(2k − 1)! sinh 2k tanh−1 x ×

dx2k+1 dx (1 − x2 )k
(2k − 1)! d
sinh 2k tanh−1 x .

+ k
×
2
(1 − x ) dx

202
Using the chain rule,
d 1 2kx
k
= k+1
dx (1 − x )
2 (1 − x2 )
and
d 2k
sinh 2k tanh−1 x = cosh 2k tanh−1 x ·
 
.
dx 1 − x2
Hence,

d2k+1 2kx
tanh−1 x =(2k − 1)! sinh 2k tanh−1 x ×

dx 2k+1 k+1
(1 − x2 )
2k cosh 2k tanh−1 x

(2k − 1)!
+ k
×
(1 − x2 ) 1 − x2

(2k)! x
sinh 2k tanh−1 x

= 1 √
k+ 1−x 2
(1 − x2 ) 2

1
cosh 2k tanh−1 x

+√
1 − x2
(2k)!  −1
 −1

= k+ 1 sinh tanh x sinh 2k tanh x
(1 − x2 ) 2
+ cosh tanh−1 x cosh 2k tanh−1 x .
 

Using the identity

cosh(x + y) = cosh x cosh y + sinh x sinh y,

we get
d2k+1 (2k)!
tanh−1 x = −1

2k+1 1 cosh (2k + 1) tanh x .
dx k+
(1 − x2 ) 2
Differentiating both sides with respect to x again using the product rule,

d2k+2 d 1
tanh−1 x =(2k)! cosh (2k + 1) tanh−1 x ×

dx 2k+2 dx (1 − x2 )k+ 21
(2k)! d
cosh (2k + 1) tanh−1 x .

+ 1 ×
k+
(1 − x2 ) 2 dx

Using the chain rule,


d 1 (2k + 1)x
=
dx (1 − x2 )k+ 12 k+ 3
(1 − x2 ) 2
and
d  2k + 1
cosh (2k + 1) tanh−1 x = sinh (2k + 1) tanh−1 x ·

.
dx 1 − x2

203
Hence,

d2k+2 (2k + 1)x


tanh−1 x =(2k)! cosh (2k + 1) tanh−1 x ×

dx 2k+2 k+ 3
(1 − x2 ) 2
(2k + 1) sinh (2k + 1) tanh−1 x

(2k)!
+ k+ 1
×
(1 − x2 ) 2 1 − x2

(2k + 1)! x
cosh (2k + 1) tanh−1 x

= k+1

2 1−x 2
(1 − x )

1
sinh (2k + 1) tanh−1 x

+√
1 − x2
(2k + 1)! 
sinh tanh−1 x cosh (2k + 1) tanh−1 x
 
= k+1
(1 − x2 )
+ cosh tanh−1 x sinh (2k + 1) tanh−1 x .
 

Using the identity

sinh(x + y) = sinh x cosh y + cosh x sinh y,

we get
d2k+2 (2k + 1)!
tanh−1 x = sinh (2k + 2) tanh−1 x .

dx 2k+2 k+1
(1 − x2 )
Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

d2n (2n − 1)!


tanh−1 x = −1

2n 2 n sinh 2n tanh x
dx (1 − x )

for all n ∈ N.

204
Problem 87
Topic: Derivative, Binomial Coefficient

Problem: Prove the Leibniz rule,


n  
dn X n (j)
(f (x)g(x)) = f (x)g (n−j) (x),
dxn j=0
j

for all n ∈ N where f (n) (x) denotes the n-th derivative of f (x).

Definition - Binomial Coefficient: For non-negative integers p and r,


  ( p!
p if r ≤ p
= r!(p−r)! .
r 0 if r > p

Solution
Base Cases
We need to prove that it is true for n = 1. By definition of derivative,

f (x + h) − f (x)
f 0 (x) = lim .
h→0 h

So,
d f (x + h)g(x + h) − f (x)g(x)
(f (x)g(x)) = lim .
dx h→0 h

205
Adding and subtracting f (x)g(x + h) in the numerator,
d
(f (x)g(x))
dx
f (x + h)g(x + h) − f (x)g(x + h) + f (x)g(x + h) − f (x)g(x)
= lim
h→0 h
f (x + h)g(x + h) − f (x)g(x + h) f (x)g(x + h) − f (x)g(x)
= lim + lim
h→0 h h→0 h
   
f (x + h) − f (x) g(x + h) − g(x)
= lim · g(x) + f (x) · lim
h→0 h h→0 h
=f 0 (x)g(x) + f (x)g 0 (x).

For the right-hand side,


1      
X 1 (j) (1−j) 1 (0) (1) 1 (1)
f (x)g (x) = f (x)g (x) + f (x)g (0) (x)
j=0
j 0 1
= f 0 (x)g(x) + f (x)g 0 (x).

So, it is true that


1  
d X 1 (j)
(f (x)g(x)) = f (x)g (1−j) (x).
dx j=0
j

Induction Step
Assume
m  
dm X m (j)
m
(f (x)g(x)) = f (x)g (m−j) (x) (87.1)
dx j=0
j

for all 1 ≤ m ≤ k for some k ≥ 1. To prove the induction step, we need to prove
k+1
dk+1 X k + 1
(f (x)g(x)) = f (j) (x)g (k+1−j) (x).
dxk+1 j=0
j

By assumption (87.1) when m = k,


k  
dk X k (j)
k
(f (x)g(x)) = f (x)g (k−j) (x).
dx j=0
j

Differentiating both sides with respect to x,


k  
dk+1 d X k (j)
(f (x)g(x)) = f (x)g (k−j) (x)
dxk+1 dx j=0 j
k  
X k d  (j) 
= f (x)g (k−j) (x) .
j=0
j dx

206
By assumption (87.1) when m = 1,

1  
d X 1 (j)
(f (x)g(x)) = f (x)g (1−j) (x) = f 0 (x)g(x) + f (x)g 0 (x).
dx j=0
j

So,
d  (j) 
f (x)g (k−j) (x) = f (j+1) (x)g (k−j) (x) + f (j) (x)g (k+1−j) (x),
dx

and we have

k  h
dk+1 X k i
k+1
(f (x)g(x)) = f (j+1) (x)g (k−j) (x) + f (j) (x)g (k+1−j) (x)
dx j=0
j
k   k  
X k (j+1) (k−j)
X k
= f (x)g (x) + f (j) (x)g (k+1−j) (x).
j=0
j j=0
j

Substituting j + 1 → j,

k   k+1
X 
X k (j+1) (k−j) k
f (x)g (x) = f (j) (x)g (k+1−j) (x).
j=0
j j=1
j−1

Hence,

dk+1
(f (x)g(x))
dxk+1
k+1   k  
X k (j) (k+1−j)
X k (j)
= f (x)g (x) + f (x)g (k+1−j) (x)
j=1
j − 1 j=0
j
  k  
k (k+1) (k+1−k−1)
X k
= f (x)g (x) + f (j) (x)g (k+1−j) (x)
k j=1
j − 1
k    
X k k (0)
+ f (j) (x)g (k+1−j) (x) + f (x)g (k+1−0) (x)
j=1
j 0
k    
X k k
=f (k+1) (x)g(x) + + f (j) (x)g (k+1−j) (x) + f (x)g (k+1) (x).
j=1
j − 1 j

In problem 73, we have calculated that


     
k k k+1
+ = .
j−1 j j

207
So,

dk+1
(f (x)g(x))
dxk+1
k  
X k + 1 (j)
=f (k+1) (x)g(x) + f (x)g (k+1−j) (x) + f (x)g (k+1) (x)
j=1
j
  k  
k + 1 (k+1) X k + 1 (j)
= f (x)g (k+1−k−1) (x) + f (x)g (k+1−j) (x)
k+1 j=1
j
 
k + 1 (0)
+ f (x)g (k+1−0) (x)
0
k+1
X k + 1 
= f (j) (x)g (k+1−j) (x).
j=0
j

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
n  
dn X n (j)
(f (x)g(x)) = f (x)g (n−j) (x)
dxn j=0
j

for all n ∈ N.

208
Problem 88
Topic: Derivative

Problem: Prove
 
n n n
d Y Y X fj0 (x)
fj (x) = fj (x) ·  
dx j=1 j=1
f (x)
j=1 j

for all n ∈ N.

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. When n = 1,
 
1 0 1 1 0
d Y f (x) Y X fj (x)
fj (x) = f10 (x) = f1 (x) · 1 = fj (x) ·  .
dx j=1 f1 (x) j=1 f
j=1 j
(x)

When n = 2, by the product rule of derivative,

2
d Y d
fj (x) = (f1 (x)f2 (x)) = f10 (x)f2 (x) + f1 (x)f20 (x)
dx j=1 dx
 
2 2
f10 (x) f20 (x) fj0 (x)
  Y X
= f1 (x)f2 (x) · + = fj (x) ·  .
f1 (x) f2 (x) j=1 j=1
fj (x)

So, it is true for n = 1 and n = 2.

209
Induction Step
Assume  
m m m 0
d Y Y X f j (x)
fj (x) = fj (x) ·   (88.1)
dx j=1 j=1
f
j=1 j
(x)

for all 1 ≤ m ≤ k for some k ≥ 2. To prove the induction step, we need to prove
 
k+1 k+1 k+1
X fj0 (x)
d Y Y
fj (x) = fj (x) ·  .
dx j=1 j=1 j=1
fj (x)

By assumption (88.1) when m = 2,

f10 (x) f20 (x)


 
d
(f1 (x)f2 (x)) = f1 (x)f2 (x) · + .
dx f1 (x) f2 (x)

k
Y
Let f1 (x) → fj (x) and f2 (x) → fk+1 (x),
j=1

 −1 
k+1 k+1 k k 0
d Y Y Y d Y f (x) 
fj (x) = fj (x) ·  fj (x) · fj (x) + k+1 .

dx j=1 j=1 j=1
dx j=1
f k+1 (x)

By assumption (88.1) when m = k,


 
k k k
d Y Y X fj0 (x)
fj (x) = fj (x) ·  .
dx j=1 j=1 j=1
fj (x)

So,
k+1
d Y
fj (x)
dx j=1
 −1   
k+1 k k k 0 0
Y Y Y X fj (x) f (x) 
= fj (x) · 

fj (x) · fj (x) ·   + k+1
f (x) fk+1 (x)

j=1 j=1 j=1 j=1 j
 
k+1 k
Y X fj0 (x) fk+1
0
(x)
= fj (x) ·  + 
j=1
f (x) fk+1 (x)
j=1 j
 
k+1
Y k+1
X fj0 (x)
= fj (x) ·  .
j=1
f (x)
j=1 j

Thus, we proved that it is true for n = k +1 if it is true for n = m for all 1 ≤ m ≤ k.

210
Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
 
n n n
d Y Y X fj0 (x)
fj (x) = fj (x) ·  
dx j=1 j=1 j=1
f j (x)

for all n ∈ N.

211
Problem 89
Topic: Matrices

Problem: If A is an idempotent matrix, then prove that An = A for all n ∈ N.

Definition - Idempotent Matrix: An idempotent matrix is a matrix A such


that A2 = A.

Solution
Base Cases
We need to prove that it is true for n = 1. The base case n = 1 is trivial since it is
always true that A1 = A.

Induction Step
Assume that Ak = A if A is an idempotent matrix for some k ≥ 1. To prove the
induction step, we need to prove that Ak+1 = A if A is an idempotent matrix. By
the assumption that Ak = A, we have
Ak+1 = AAk = AA = A2 .
If A is an idempotent matrix, then by definition of idempotent matrix, A2 = A.
So, Ak+1 = A2 = A if A is an idempotent matrix. Thus, we proved that it is true
for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction, if A is an idempotent matrix,
then An = A for all n ∈ N.

212
Problem 90
Topic: Matrices

n
Problem: Let A be an invertible matrix. Prove
 that the inverse of A is equal
−1 n −1 −1 n
to the n-th power of A , i.e. (A ) = A , for all n ∈ N.

Definition - Inverse of Matrix: For an invertible matrix A, the inverse of


A is the matrix A−1 such that

AA−1 = A−1 A = I,

where I is the identity matrix. In this problem, we will use the property

(AB)−1 = B −1 A−1 .

We can check that this is true by checking that

AB B −1 A−1 = B −1 A−1 AB = I.
 

However, note that this is NOT a rigorous proof of the property.

Solution
Base Cases
We need to prove that it is true for n = 1. Since
−1 1
A1 = A−1 = A−1 ,

it is true for n = 1.

213
Induction Step
Assume −1 k
Ak = A−1 (90.1)
for some k ≥ 1. To prove the induction step, we need to prove
−1 k+1
Ak+1 = A−1 .

By using the assumption (90.1) and the property (AB)−1 = B −1 A−1 ,


−1 −1
Ak+1 = AAk
−1 −1
= Ak A
k
= A−1 A−1

k+1
= A−1 .

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
−1 n
(An ) = A−1

for all n ∈ N.

214
Problem 91
Topic: Matrices

Problem: Let A be a diagonalizable matrix, i.e. A = P DP −1 where D is a


diagonal matrix. Then, prove that An = P Dn P −1 for all n ∈ N.

Definition - Inverse of Matrix: For an invertible matrix A, the inverse of


A is the matrix A−1 such that

AA−1 = A−1 A = I,

where I is the identity matrix.

Solution
Base Cases
We need to prove that it is true for n = 1. Since

A1 = A = P DP −1 = P D1 P −1 ,

it is true for n = 1.

Induction Step
Assume
Ak = P Dk P −1 (91.1)
for some k ≥ 1. To prove the induction step, we need to prove

Ak+1 = P Dk+1 P −1 .

216
By assumption (91.1),

Ak+1 = Ak A
= P Dk P −1 P DP −1
 

= P Dk P −1 P DP −1


= P Dk IDP −1
= P Dk DP −1
= P Dk+1 P −1 .

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,

An = P Dn P −1

for all n ∈ N.

217
Problem 92
Topic: Set Theory

Problem: If Sn is a set with n elements, then prove that Sn has 2n subsets.

Definition - Subset: For two sets A and B, A is a subset of B, denoted as


A ⊆ B, when all elements in A are in B. For example, {1, 3, 5} ⊆ {1, 2, 3, 5, 6}.
Formally,
A ⊆ B ⇔ [∀x, x ∈ A ⇒ x ∈ B].
Also, the empty set ∅ is a subset of any set, but this needs to be proven using
the formal definition of subset. However, we will not prove this here.

Definition - Set Union: For two sets A and B, the union of A and B,
denoted as A ∪ B, is the set containing all elements in A and all elements in
B. For example, {1, 2, 3} ∪ {1, 3, 5} = {1, 2, 3, 5}. Formally,

x ∈ A ∪ B ⇔ [x ∈ A or x ∈ B].

Solution
Base Cases
We need to prove that it is true for n = 1. Let S1 be a set with 1 element,

S1 = {a1 }.

Then, it has 2 subsets, ∅ and {a1 }. So, it is true that a set with 1 element S1 has
21 subsets.

218
Induction Step
Assume that a k-element set Sk ,

Sk = {a1 , a2 , · · · , ak },

has 2k subsets. To prove the induction step, we need to prove that a k + 1-element
set Sk+1 ,
Sk+1 = {a1 , a2 , · · · , ak , ak+1 },
has 2k+1 subsets. Note that

Sk+1 = {a1 , a2 , · · · , ak } ∪ {ak+1 } = Sk ∪ {ak+1 }.

A subset of Sk+1 either contains or does not contain ak+1 .

Subsets of Sk+1 that do not contain ak+1 are subsets of Sk . By the assumption
that Sk has 2k subsets, Sk+1 has 2k subsets that do not contain ak+1 .

Subsets of Sk+1 that contain ak+1 are the unions of a subset of Sk with {ak+1 }.
Since there are 2k subsets of Sk by assumption, there are 2k sets formed by taking
the union of a subset of Sk with {ak+1 }. So, Sk+1 has 2k subsets that contain ak+1 .

Hence, Sk+1 has 2k + 2k = 2k+1 subsets. Thus, we proved that it is true for
n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction, Sn has 2n subsets if it has n
elements for all n ∈ N.

219
Problem 93
Topic: Set Theory

Problem: Let S1 , S2 , · · · , Sn be sets such that Sn ⊆ Sn−1 ⊆ · · · ⊆ S2 ⊆ S1 ,


then prove that
n
\
Si = Sn
i=1

for all n ∈ N.

Definition - Set Intersection: For two sets A and B, the intersection of A


and B, denoted as A ∩ B, is the set containing all elements that are in both A
and B. For example, {1, 2, 3} ∩ {1, 3, 5} = {1, 3}. Formally,

x ∈ A ∩ B ⇔ [x ∈ A and x ∈ B].

For more than 2 sets, we have


n
\
Ai = A1 ∩ A2 ∩ · · · ∩ An .
i=1

Definition - Subset: For two sets A and B, A is a subset of B, denoted as


A ⊆ B, when all elements in A are in B. For example, {1, 3, 5} ⊆ {1, 2, 3, 5, 6}.
Formally,
A ⊆ B ⇔ [∀x, x ∈ A ⇒ x ∈ B].

Definition - Equal Sets: For two sets A and B, A is equal to B, denoted as


A = B, when all elements in A are in B and all elements in B are in A. For

220
example, {1, 3, 5} = {1, 1, 3, 5, 5}. Formally,

A = B ⇔ [∀x, x ∈ A ⇔ x ∈ B],

or equivalently
A = B ⇔ [A ⊆ B and B ⊆ A].

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. The case when n = 1 is
trivial since it is always true that
1
\
Si = S1 .
i=1

For n = 2, we need to prove that


2
\
Si = S1 ∩ S2 = S2
i=1

if S2 ⊆ S1 .

First, we need to prove S1 ∩ S2 ⊆ S2 . Let x ∈ S1 ∩ S2 . By definition of set


intersection,
x ∈ S1 ∩ S2 ⇒ [x ∈ S1 and x ∈ S2 ].
Since the statement “x ∈ S1 and x ∈ S2 ” is true only when “x ∈ S1 ” and “x ∈ S2 ”
are both true,
[x ∈ S1 and x ∈ S2 ] ⇒ x ∈ S2 .
Hence,
∀x, x ∈ S1 ∩ S2 ⇒ x ∈ S2 ,
and S1 ∩ S2 ⊆ S2 by definition of subset.

Next, we need to prove S2 ⊆ S1 ∩ S2 . Let x ∈ S2 . If S2 ⊆ S1 , then by defi-


nition of subset,
x ∈ S2 ⇒ x ∈ S1 .
So,
x ∈ S2 ⇒ [x ∈ S1 and x ∈ S2 ].
By definition of set intersection,
[x ∈ S1 and x ∈ S2 ] ⇒ x ∈ S1 ∩ S2 .

221
Hence,
∀x, x ∈ S2 ⇒ x ∈ S1 ∩ S2 ,
and S2 ⊆ S1 ∩ S2 by definition of subset.

Thus, S1 ∩ S2 ⊆ S2 and S2 ⊆ S1 ∩ S2 . By definition of equal sets, S1 ∩ S2 = S2 if


S2 ⊆ S1 .

Induction Step
Assume
m
\
Si = Sm (93.1)
i=1
if Sm ⊆ Sm−1 ⊆ · · · ⊆ S2 ⊆ S1 for all 1 ≤ m ≤ k for some k ≥ 2. To prove the
induction step, we need to prove
k+1
\
Si = Sk+1
i=1

if Sk+1 ⊆ Sk ⊆ · · · ⊆ S2 ⊆ S1 . By the assumption (93.1) when m = k,


k
\
Si = Sk
i=1

if Sk ⊆ Sk−1 ⊆ · · · ⊆ S2 ⊆ S1 . So,
k+1
\ k
\
Si = Si ∩ Sk+1 = Sk ∩ Sk+1 .
i=1 i=1

By the assumption (93.1) when m = 2, S1 ∩ S2 = S2 if S2 ⊆ S1 . Let S1 → Sk and


S2 → Sk+1 ,
Sk ∩ Sk+1 = Sk+1
if Sk+1 ⊆ Sk . Hence,
k+1
\
Si = Sk+1
i=1
if Sk+1 ⊆ Sk ⊆ · · · ⊆ S2 ⊆ S1 . Thus, we proved that it is true for n = k + 1 if it is
true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
n
\
Si = Sn
i=1

if Sn ⊆ Sn−1 ⊆ · · · ⊆ S2 ⊆ S1 for all n ∈ N.

222
Problem 94
Topic: Set Theory

Problem: Let S1 , S2 , · · · , Sn be sets such that Sn ⊆ Sn−1 ⊆ · · · ⊆ S2 ⊆ S1 ,


then prove that
n
[
Si = S1
i=1

for all n ∈ N.

Definition - Set Union: For two sets A and B, the union of A and B,
denoted as A ∪ B, is the set containing all elements in A and all elements in
B. For example, {1, 2, 3} ∪ {1, 3, 5} = {1, 2, 3, 5}. Formally,

x ∈ A ∪ B ⇔ [x ∈ A or x ∈ B].

For more than 2 sets, we have


n
[
Ai = A1 ∪ A2 ∪ · · · ∪ An .
i=1

Definition - Subset: For two sets A and B, A is a subset of B, denoted as


A ⊆ B, when all elements in A are in B. For example, {1, 3, 5} ⊆ {1, 2, 3, 5, 6}.
Formally,
A ⊆ B ⇔ [∀x, x ∈ A ⇒ x ∈ B].

Definition - Equal Sets: For two sets A and B, A is equal to B, denoted as


A = B, when all elements in A are in B and all elements in B are in A. For

223
example, {1, 3, 5} = {1, 1, 3, 5, 5}. Formally,

A = B ⇔ [∀x, x ∈ A ⇔ x ∈ B],

or equivalently
A = B ⇔ [A ⊆ B and B ⊆ A].

Solution
Base Cases
We need to prove that it is true for n = 1 and n = 2. The case when n = 1 is
trivial since it is always true that
1
[
Si = S1 .
i=1

For n = 2, we need to prove that


2
[
Si = S1 ∪ S2 = S1
i=1

if S2 ⊆ S1 .

First, we need to prove S1 ∪ S2 ⊆ S1 . Let x ∈ S1 ∪ S2 . By definition of set


union,
x ∈ S1 ∪ S2 ⇒ [x ∈ S1 or x ∈ S2 ].
If S2 ⊆ S1 , then by definition of subset,

x ∈ S2 ⇒ x ∈ S1 .

So,
[x ∈ S1 or x ∈ S2 ] ⇒ [x ∈ S1 or x ∈ S1 ] ⇒ x ∈ S1 .
Hence,
∀x, x ∈ S1 ∪ S2 ⇒ x ∈ S1 ,
and S1 ∪ S2 ⊆ S1 by definition of subset.

Next, we need to prove S1 ⊆ S1 ∪ S2 . Let x ∈ S1 . Since the statement “x ∈ S1 or


x ∈ S2 ” is true when either “x ∈ S1 ” or “x ∈ S2 ” is true,

x ∈ S1 ⇒ [x ∈ S1 or x ∈ S2 ].

224
By definition of set union,
[x ∈ S1 or x ∈ S2 ] ⇒ x ∈ S1 ∪ S2 .
Hence,
∀x, x ∈ S1 ⇒ x ∈ S1 ∪ S2 ,
and S1 ⊆ S1 ∪ S2 by definition of subset.

Thus, S1 ∪ S2 ⊆ S1 and S1 ⊆ S1 ∪ S2 . By definition of equal sets, S1 ∪ S2 = S1 if


S2 ⊆ S1 .

Induction Step
Assume
m
[
Si = S1 (94.1)
i=1
if Sm ⊆ Sm−1 ⊆ · · · ⊆ S2 ⊆ S1 for all 1 ≤ m ≤ k for some k ≥ 2. To prove the
induction step, we need to prove
k+1
[
Si = S1
i=1

if Sk+1 ⊆ Sk ⊆ · · · ⊆ S2 ⊆ S1 . By the assumption (94.1) when m = k,


k
[
Si = S1
i=1

if Sk ⊆ Sk−1 ⊆ · · · ⊆ S2 ⊆ S1 . So,
k+1
[ k
[
Si = Si ∪ Sk+1 = S1 ∪ Sk+1 .
i=1 i=1

By the assumption (94.1) when m = 2, S1 ∪ S2 = S1 if S2 ⊆ S1 . Let S2 → Sk+1 ,


S1 ∪ Sk+1 = S1
if Sk+1 ⊆ S1 . Hence,
k+1
[
Si = S1
i=1
if Sk+1 ⊆ Sk ⊆ · · · ⊆ S2 ⊆ S1 . Thus, we proved that it is true for n = k + 1 if it is
true for n = m for all 1 ≤ m ≤ k.

Conclusion
We proved that it is true for n = 1 and n = 2. Then, we proved that it is true for
n = k + 1 if it is true for n = m for all 1 ≤ m ≤ k. Therefore, by strong induction,
n
[
Si = S1
i=1

if Sn ⊆ Sn−1 ⊆ · · · ⊆ S2 ⊆ S1 for all n ∈ N.

225
Problem 95
Topic: Integrals

Problem: Prove Z ∞
xn e−x dx = n!.
0

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. By integration by parts,
Z ∞ Z ∞
−x −x ∞
e−x dx
 
xe dx = −xe 0
+
0 0
Z ∞
= e−x dx
0 ∞
= −e−x 0 = 1 = 1!,
so it is true for n = 1.

Induction Step
Assume Z ∞
xk e−x dx = k! (95.1)
0

226
for some k ≥ 1. To prove the induction step, we need to prove
Z ∞
xk+1 e−x dx = (k + 1)!.
0

By integration by parts,
Z ∞ Z ∞
∞
xk+1 e−x dx = −xk+1 e−x 0 + (k + 1) xk e−x dx

0 0
Z ∞
= (k + 1) xk e−x dx.
0

Using the assumption (95.1),


Z ∞
xk+1 e−x dx = (k + 1) · k! = (k + 1)!.
0

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
Z ∞
xn e−x dx = n!
0

for all n ∈ N.

227
Problem 96
Topic: Integrals, Trigonometry

Problem: Prove
Z π4 n
(−1)n ln 2 (−1)n X (−1)j
tan2n+1 x dx = + .
0 2 2 j=1 j

Solution
Base Cases

We need to prove that it is true for n = 1. By using the identity 1+tan2 x = sec2 x,

Z π Z π
4 4
tan3 x dx = tan x sec2 x − 1 dx

0 0
Z π Z π
4 4
= tan x sec2 x dx − tan x dx
0 0
 π4
tan2 x

= + ln (cos x)
2 0
1 ln 2
= −
2 2
1
(−1)1 ln 2 (−1)1 X (−1)j
= + ,
2 2 j=1 j

so it is true for n = 1.

228
Induction Step
Assume
π k
(−1)k ln 2 (−1)k X (−1)j
Z 4
tan2k+1 x dx = + (96.1)
0 2 2 j=1 j

for some k ≥ 1. To prove the induction step, we need to prove


π k+1
(−1)k+1 ln 2 (−1)k+1 X (−1)j
Z 4
tan2k+3 x dx = + .
0 2 2 j=1
j

By using the identity 1 + tan2 x = sec2 x,


Z π Z π
4 4
2k+3
tan2k+1 x sec2 x − 1 dx

tan x dx =
0 0
Z π Z π
4 4
= tan2k+1 x sec2 x dx − tan2k+1 x dx
0 0
π Z π
tan2k+2 x 4

4
= − tan2k+1 x dx
2k + 2 0 0
Z π4
1
= − tan2k+1 x dx.
2k + 2 0

By the assumption (96.1),


π k
(−1)k ln 2 (−1)k X (−1)j
Z 4 1
tan2k+3 x dx = − −
0 2k + 2 2 2 j=1 j
k
(−1)k+1 (−1)k+1 (−1)k+1 ln 2 (−1)k+1 X (−1)j
= + +
2(k + 1) 2 2 j=1
j
k+1
(−1)k+1 ln 2 (−1)k+1 X (−1)j
= + .
2 2 j=1
j

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
π n
(−1)n ln 2 (−1)n X (−1)j
Z 4
tan2n+1 x dx = +
0 2 2 j=1 j

for all n ∈ N.

229
Problem 97
Topic: Integrals, Trigonometry

Problem: Prove Z π
2 (2n)! π
sin2n x dx = · .
0 4n (n!)2 2

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
1−cos (2x)
We need to prove that it is true for n = 1. By using the identity sin2 x = 2 ,

π π
1 − cos (2x)
Z 2
Z 2
sin2 x dx = dx
0 0 2
 π
x sin (2x) 2
= −
2 4 0
π 2! π
= = 1 · ,
4 4 (1!)2 2

so it is true for n = 1.

230
Induction Step
Assume
Z π
2 (2k)! π
sin2k x dx = · (97.1)
0 4k (k!)2 2

for some k ≥ 1. To prove the induction step, we need to prove


Z π
2 (2k + 2)! π
sin2k+2 x dx = · .
0 4k+1 ((k 2
+ 1)!) 2

By using the identity sin2 x + cos2 x = 1,


Z π Z π
2 2
2k+2
sin2k x 1 − cos2 x dx

sin x dx =
0 0
Z π Z π
2 2
2k
= sin x dx − sin2k x cos2 x dx. (97.2)
0 0

By integration by parts,

π
" # π2 Z π2
sin2k+1 x
Z 2
2k 1
sin 2
x cos x dx = · cos x + sin2k+2 x dx
0 2k + 1 2k + 1 0
0
Z π2
1
= sin2k+2 x dx.
2k + 1 0

Substituting this into (97.2),


Z π Z π Z π
2
2k+2
2
2k 1 2
sin x dx = sin x dx − sin2k+2 x dx
0 0 2k + 1 0
Z π Z π
2k + 2 2
2k+2
2
sin x dx = sin2k x dx
2k + 1 0 0
Z π Z π
2 2k + 1 2
sin2k+2 x dx = sin2k x dx.
0 2k + 2 0

By the assumption (97.1),


Z π
2 2k + 1 (2k)! π
sin2k+2 x dx = · ·
0 2k + 2 4k (k!)2 2
2k + 2 2k + 1 (2k)! π
= · · ·
2k + 2 2k + 2 4k (k!)2 2
(2k + 2)! π
= k+1 · .
4 ((k + 1)!)2 2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

231
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
Z π
2 (2n)! π
sin2n x dx = ·
0 4n (n!)2 2

for all n ∈ N.

232
Problem 98
Topic: Integrals, Trigonometry

Problem: Prove
 
π n
(2n)!  π X 2j (j!)2 
Z 4
2n
sin x dx = n − .
0 4 (n!)2 4 j=1 2j(2j)!

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
1−cos (2x)
We need to prove that it is true for n = 1. By using the identity sin2 x = 2 ,
Z π4 Z π4
2 1 − cos (2x)
sin x dx = dx
0 0 2
 π
x sin (2x) 4
= −
2 4 0
π 1
= −
8 4  
1
2!  π X 2j (j!)2 
= − ,
41 (1!)2 4 j=1 2j(2j)!

233
so it is true for n = 1.

Induction Step
Assume
 
Z π k j 2
4 (2k)! π X 2 (j!)
sin2k x dx = k  −  (98.1)
0 4 (k!)2 4 j=1 2j(2j)!

for some k ≥ 1. To prove the induction step, we need to prove


 
Z π k+1 j 2
4 (2k + 2)! π − 2 (j!)  X
sin2k+2 x dx = .
0 4k+1 ((k + 1)!)2 4 j=1 2j(2j)!

By using the identity sin2 x + cos2 x = 1,

Z π Z π
4 4
sin2k+2 x dx = sin2k x 1 − cos2 x dx

0 0
Z π Z π
4 4
= sin2k x dx − sin2k x cos2 x dx. (98.2)
0 0

By integration by parts,

π
" # π4 Z π4
sin2k+1 x
Z 4
2k 1
sin x cos x dx =2
· cos x + sin2k+2 x dx
0 2k + 1 2k + 1 0
0
Z π4
1 1
= + sin2k+2 x dx.
2k+1 (2k + 1) 2k + 1 0

Substituting this into (98.2),

Z π Z π Z π
4 4 1 1 4
sin2k+2 x dx = sin2k x dx − − sin2k+2 x dx.
0 0 2k+1 (2k + 1) 2k + 1 0

So,

Z π Z π
2k + 2 4 4 1
sin2k+2 x dx = sin2k x dx −
2k + 1 0 0 2k+1 (2k + 1)
Z π Z π
4 2k + 1 4 1
sin2k+2 x dx = sin2k x dx − .
0 2k + 2 0 2k+2 (k + 1)

234
By the assumption (98.1),
 
Z π k j 2
4 2k + 1 (2k)!  π X 2 (j!)  1
sin2k+2 x dx = · − − k+2
0 2k + 2 4k (k!)2 4 j=1 2j(2j)! 2 (k + 1)
 
k j 2
(2k + 2)! π −
X 2 (j!)  1
= k+1 − k+2
4 ((k + 1)!)2 4 j=1 2j(2j)! 2 (k + 1)
 
k+1 j 2
(2k + 2)! π −
X 2 (j!) 
= k+1 .
4 ((k + 1)!)2 4 j=1 2j(2j)!

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 
Z π4 n j 2
(2n)! π X 2 (j!)
sin2n x dx = n  − 
0 4 (n!)2 4 j=1 2j(2j)!

for all n ∈ N.

235
Problem 99
Topic: Integrals, Trigonometry

Problem: Prove
π n−1
22n−1 (n!)2 X (2j)!
Z 4
sec2n x dx = .
0 n (2n)! j=0 2j (j!)2

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. Since
π 1−1
22·1−1 (1!)2 X (2j)!
Z 4 π
sec2 x dx = [tan x]04 = 1 = ,
0 1 · 2! j=0
2j (j!)2

it is true for n = 1.

Induction Step
Assume
π k−1
22k−1 (k!)2 X (2j)!
Z 4
2k
sec x dx = (99.1)
0 k (2k)! j=0 2j (j!)2

236
for some k ≥ 1. To prove the induction step, we need to prove
π k
22k+1 ((k + 1)!)2 X (2j)!
Z 4
sec2k+2 x dx = .
0 (k + 1)(2k + 2)! j=0 2j (j!)2

By integration by parts and using the identity 1 + tan2 x = sec2 x,


Z π Z π
4 4
sec2k+2 x dx = sec2k x sec2 x dx
0 0
Z π
π 4
= sec2k x tan x 04 − 2k sec2k x tan2 x dx

0
Z π
4
k
sec2k x sec2 x − 1 dx

= 2 − 2k
0
Z π Z π
4 4
= 2k − 2k sec2k+2 x dx + 2k sec2k x dx.
0 0

So,
π π
2k
Z Z
4 2k 4
sec2k+2 x dx = + sec2k x dx.
0 2k + 1 2k + 1 0

By the assumption (99.1),


π k−1
2k 22k−1 (k!)2 X (2j)!
Z 4 2k
sec2k+2 x dx = + ·
0 2k + 1 2k + 1 k (2k)! j=0 2j (j!)2
k−1
2k 22k+1 ((k + 1)!)2 X (2j)!
= +
2k + 1 (k + 1)(2k + 2)! j=0 2j (j!)2
k
22k+1 ((k + 1)!)2 X (2j)!
= .
(k + 1)(2k + 2)! j=0 2j (j!)2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
π n−1
22n−1 (n!)2 X (2j)!
Z 4
sec2n x dx =
0 n (2n)! j=0 2j (j!)2

for all n ∈ N.

237
Problem 100
Topic: Integrals, Trigonometry

Problem: Prove
π n
(−1)n (2n)!  π 2j
Z 2 X
x2n cos x dx = .
0 j=0
(−1)j (2j)! 2

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
We need to prove that it is true for n = 1. By integration by parts twice,
Z π2 Z π2
2
 2  π2
x cos x dx = x sin x 0 − 2 x sin x dx
0 0
Z π
 π 2 2
= −2 x sin x dx
2 0
Z π
 π 2 π 2
= + [2x cos x]02 − 2 cos x dx
2 0
Z π
 π 2 2  π 2
= −2 cos x dx = − 2.
2 0 2

238
Since
1
X (−1)1 2!  π 2j  π 2
= − 2,
j=0
(−1)j (2j)! 2 2

it is true that
π 1
(−1)1 2!  π 2j
Z 2 X
x2 cos x dx = .
0 j=0
(−1)j (2j)! 2

Induction Step
Assume
π k
(−1)k (2k)!  π 2j
Z 2 X
x2k cos x dx = (100.1)
0 j=0
(−1)j (2j)! 2

for some k ≥ 1. To prove the induction step, we need to prove


π k+1
(−1)k+1 (2k + 2)!  π 2j
Z 2 X
2k+2
x cos x dx = .
0 j=0
(−1)j (2j)! 2

By integration by parts twice,


Z π
2
x2k+2 cos x dx
0
Z π
π 2
= x2k+2 sin x 02 − (2k + 2) x2k+1 sin x dx

0
Z π
 π 2k+2 2
= − (2k + 2) x2k+1 sin x dx
2 0
Z π
 π 2k+2 π 2
+ (2k + 2) x2k+1 cos x 02 − (2k + 2)(2k + 1) x2k cos x dx

=
2 0
Z π
 π 2k+2 2
= − (2k + 2)(2k + 1) x2k cos x dx.
2 0

By the assumption (100.1),


π k
(−1)k (2k)!  π 2j
Z 2  π 2k+2 X
x2k+2 cos x dx = − (2k + 2)(2k + 1)
0 2 j=0
(−1)j (2j)! 2
k
 π 2k+2 X (−1)k+1 (2k + 2)!  π 2j
= +
2 j=0
(−1)j (2j)! 2
k+1
X (−1)k+1 (2k + 2)!  π 2j
= .
j=0
(−1)j (2j)! 2

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

239
Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
π n
(−1)n (2n)!  π 2j
Z 2 X
x2n cos x dx =
0 j=0
(−1)j (2j)! 2

for all n ∈ N.

240
Problem 101
Topic: Integrals, Trigonometry

Problem: Prove
 
π n−1
(2n)!  π 2 X 4j (j!)2 
Z 2
2n
x cos x dx = n − .
0 4 (n!)2 8 j=0
(2j + 2)!

Definition - Factorial:

n! = n · (n − 1) · (n − 2) · · · 2 · 1.

0! = 1.

Solution
Base Cases
1+cos (2x)
We need to prove that it is true for n = 1. By using the identity cos2 x = 2 ,
Z π2 Z π2
x + x cos (2x)
x cos2 x dx = dx
0 0 2
 2 π
x x sin (2x) cos (2x) 2
= + +
4 4 8 0
π2 1
= −
16 4  
1−1
2!  π 2 X 4j (j!)2 
= 1 − ,
4 (1!)2 8 j=0
(2j + 2)!

241
so it is true for n = 1.

Induction Step

Assume
 
Z π
2 k−1 j 2
2 (2k)!  π 4 (j!)  X
x cos2k x dx = − (101.1)
0 4k (k!)2 8 j=0
(2j + 2)!

for some k ≥ 1. To prove the induction step, we need to prove

 
Z π k j 2 2
2 (2k + 2)! π −
X 4 (j!) 
x cos2k+2 x dx = k+1 .
0 4 ((k + 1)!)2 8 j=0
(2j + 2)!

By using the identity sin2 x + cos2 x = 1,

Z π Z π
2 2
2k+2
x cos2k x 1 − sin2 x dx

x cos x dx =
0 0
Z π Z π
2 2
= x cos 2k
x dx − x cos2k x sin2 x dx. (101.2)
0 0

By integration by parts,

Z π
2
− x cos2k x sin2 x dx
0
Z π
2
(x sin x) − cos2k x sin x dx

=
0
π Z π2
cos2k+1 x 2

1
= x sin x · − (sin x + x cos x) cos2k+1 x dx
2k + 1 0 2k + 1 0
Z π2 Z π2
1 2k+1 1
=− sin x cos x dx − x cos2k+2 x dx
2k + 1 0 2k + 1 0
Z π2
1 1
=− − x cos2k+2 x dx.
(2k + 2)(2k + 1) 2k + 1 0

Substituting this into (101.2) and simplifying algebraically gives

Z π Z π
2 2k + 1 2 1
x cos2k+2 x dx = x cos2k x dx − .
0 2k + 2 0 4(k + 1)2

242
By the assumption (101.1),
 
Z π
2 k−1 j 2
2 2k + 1 (2k)!  π X 4 (j!)  1
x cos2k+2 x dx = · − −
0 2k + 2 4k (k!)2 8 j=0
(2j + 2)! 4(k + 1)2
 
2 k−1 j 2
(2k + 2)! π −
X 4 (j!)  1
= k+1 −
4 ((k + 1)!)2 8 j=0
(2j + 2)! 4(k + 1)2
 
2 k j 2
(2k + 2)! π −
X 4 (j!) 
= k+1 .
4 ((k + 1)!)2 8 j=0
(2j + 2)!

Thus, we proved that it is true for n = k + 1 if it is true for n = k.

Conclusion
We proved that it is true for n = 1. Then, we proved that it is true for n = k + 1
if it is true for n = k. Therefore, by weak induction,
 
Z π2 2 n−1
X 4j (j!)2
(2n)! π
x cos2n x dx = n  − 
0 4 (n!)2 8 j=0
(2j + 2)!

for all n ∈ N.

243
Acknowledgements

In this section, I would like to mention some people whom I would like to thank
for their help during the process of producing this book.

On Instagram, I have established an edit team consisting of a few members with


various math backgrounds. I would like to thank those members for their helpful
feedback and suggestions for the book. The following is the list of the members of
the edit team in no particular order:

• Thang Pang Ern


• Keyvon Rashidi
• Mislav Plavac

I also appreciate Andrzej Kukla, the cover designer of this book, and Vy Nguyen
Tong, the illustrator of this book, for their artistic contribution to make this book
more than just a boring math book with a simple cover and a long list of problems.

245
Reviews

“In the ideal situation, every college student should learn some mathematics, with
the depth and difficulty largely varying. As in the practice of most refined uni-
versities, mathematical education should never be standardized. Meanwhile, math
induction is one of the key tools that students need to know. Based on my teaching
experiences to many Chinese and international students, I found Tran’s book fairly
interesting and accessible.

Mathematical Induction 101 contains 101 carefully chosen exercise problems. I


believe that Tran’s book will prove to be useful for college students in the broad
range of arts and sciences. Hopefully many will appreciate the beauty and power
of mathematical reasoning with mastery of math induction after using this book.”

Dr. Chunwei Song


Professor of Mathematics, Peking University

“This book provides an interesting collection of identities and inequalities that


can be proved by mathematical induction. These 101 practices will be useful to
learn standard techniques. Some of the topics and their solutions would also be
found stimulating even for advanced students of mathematics.”

Dr. Hiroaki Nakamura


Professor of Mathematics, Osaka University

“This book provides almost all types of problems in high school and collegiate
mathematics which can be solved by mathematical induction. To each problem,
a self-contained and detailed solution is given. Difficulty ranges from quite ele-
mentary to somewhat complex. Among them there are many interesting ones.
I am particularly interested in problems related to the Fibonacci sequence, floor
function, and iterated functions.”

Dr. Young-One Kim


Professor Emeritus of Mathematics, Seoul National University

246
“As a mathematics student, this is a book I would have loved in my introductory
proof class. Duc Van Khanh Tran has written up a fantastic way to both learn
and practice one of the most essential tools in a mathematician’s toolkit: Induction!

It’s difficult to find such a large, comprehensive array of well-written and demon-
strative practice problems, which makes this all the more valuable a resource. Duc
Van Khanh Tran carefully pens each solution to be as intuitive and explanatory as
possible, so the risk of confusion at a solution is minimal.

The vast amount of problems is complemented by the vast array of topics the
book covers. Inequalities, sequences, trigonometry, power series, derivatives, and
integrals are only some of the concepts found, guaranteeing that anyone interested
in math will get something new out of this book. I’ve tried a few of these for fun!
I really like the integral problems myself.

All in all, if you’d like to learn about induction, hone your induction skills, or
just want a bunch of fun induction problems, purchase away! You won’t regret it!”

Said Kaili
Undergraduate Student of Mathematics, University of Virginia

“Despite minimal prior proof background, I can grasp the main scopes of the book
without much difficulty. Indeed, Duc Tran’s book provides clear instructions, com-
prehensive content, and helpful problems that ease students into learning mathe-
matical induction.”
Ky Minh Vinh Nguyen
Undergraduate Student of Informatics, University of Texas at Austin

247

You might also like