1.

Let T be the linear transformation such that: 0 / 1 point

⎛ ⎡ 1⎤ ⎞ ⎡ 1⎤
T ⎜ ⎢ 0⎥ ⎟ = ⎢ 3⎥
​ ​ ​ ​ ​ ​ ​ ​

⎝⎣0⎦⎠ ⎣−4⎦
⎛ ⎡ 0⎤ ⎞ ⎡ 2⎤
T ⎜⎢1⎥⎟ = ⎢−1⎥
​ ​ ​ ​ ​ ​ ​ ​

⎝⎣0⎦⎠ ⎣−3⎦
⎛ ⎡ 0⎤ ⎞ ⎡ 4 ⎤
T ⎜ ⎢ 0⎥ ⎟ = ⎢ 5 ⎥
​ ​ ​ ​ ​ ​ ​ ​

⎝⎣1⎦⎠ ⎣−11⎦
Find its rank.

Incorrect
Remember to check if all image vectors are linearly independent or dependent. You may want to review the lectures on
Matrix row-reduction and Singularity and rank of linear transformations.

2. Let M be a square matrix. 1 / 1 point

Check all that are true.

1 0
The determinant is the area of a parallelogram spanned by M and the vectors [ ] and [ ]. Therefore, it is always positive.
0 1
​ ​

n n
If det(M ) = 5, then det(M ) = 5 .

Correct
This is a straightforward application of the product rule for determinants:
det(M n ) = det(M ⋅ M ⋯ M ) = det(M ) ⋅ det(M ) ⋯ det(M ) = 5n

If M has size n, then it has n distinct eigenvalues.

If M is non-singular, then so is M −1 .

Correct
−1 1
If M is non-singular, then det(M ) =
 0. Since det(M ) =
​

det(M )
, then det(M −1 ) ​
=
 0.
​

3. Let 1 / 1 point

⎡ 1 0 1⎤
M = ⎢ 0 1 0⎥ ​ ​ ​ ​ ​

⎣ 1 1 1⎦
⎡ 2 8 7⎤
N = ⎢ 4 3 9⎥ ​ ​ ​

⎣ 1 9 5⎦
The value for det(M ⋅ N ) is:

Correct
Note that the third row of M is the sum of the first two, therefore det(M ) = 0. Therefore
det(M ⋅ N ) = det(M ) ⋅ det(N ) = 0 ⋅ det(N ) = 0.

4. What is the span of the following vectors vectors? 1 / 1 point

⎡ ⎤ ⎡ ⎤ ⎡ ⎤
 ⎡ 2⎤ ⎡ 1⎤ ⎡ 1⎤
⎢ 1⎥ , ​ ​ ​

⎢ 0⎥ , ​ ​ ​

⎢ 0⎥​ ​ ​

⎣ 1⎦ ⎣ 2⎦ ⎣ 1⎦

The entire 3 dimensional space.

A plane in a 3 dimensional space

Correct
That is correct. Because the three vectors are linearly independent, then they span all the space. Note that this also makes
them a basis.

5. Select all the options that are a basis for the 3D space 0 / 1 point

⎡ 2⎤ ⎡0⎤ ⎡0⎤
⎢ 3⎥ ,
​ ​ ​

⎢2.5⎥ ,
​ ​ ​

⎢5.5⎥​ ​ ​

⎣ 1⎦ ⎣3.5⎦ ⎣4.5⎦

⎡−1⎤ ⎡ 1⎤ ⎡ 1⎤
⎢ 2 ⎥, ​ ​ ​

⎢ 2⎥ , ​ ​ ​

⎢ 0⎥ ​ ​ ​

⎣1⎦ ⎣ 2⎦ ⎣ 1⎦

Correct
That is correct, since the three vectors are linearly independent, they form a basis for 3D space.

⎡ 1⎤ ⎡ 0⎤ ⎡ 1⎤
⎢ 0⎥ , ​ ​ ​

⎢ 1⎥ ,
​ ​ ​

⎢ 0⎥ ​ ​ ​

⎣ 1⎦ ⎣ 0⎦ ⎣ 0⎦

Correct
That is correct, since the three vectors are linearly independent, they form a basis for 3D space.

⎡ 2⎤ ⎡0⎤ ⎡−1⎤
⎢ 3⎥ ,
​ ​ ​

⎢−1⎥ , ​ ​ ​

⎢1⎥ ​ ​ ​

⎣ 1⎦ ⎣2⎦ ⎣4⎦

⎡−1⎤ ⎡ 1⎤ ⎡ 1⎤
⎢ 2 ⎥, ​ ​ ​

⎢ 2⎥ , ​ ​ ​

⎢ 6⎥ ​ ​ ​

⎣1⎦ ⎣ 2⎦ ⎣ 5⎦

You didn’t select all the correct answers

6. Select the characteristic polynomial for the given matrix. 1 / 1 point

2 1
M =[ ]
−3 6
​ ​

λ3 − 8λ + 15

λ2 + 8λ + 15

λ2 − 8λ − 1
 λ2 − 8λ + 15

Correct
The chacacteristic polynomial is the polynomial corresponding to the following:
det (M − λI )
Where
1 0 λ 0
λI = λ ⋅ [ ]=[ ]
0 1 0 λ
​ ​ ​ ​

Therefore,
2 1 λ 0 2−λ 1
det (M − λI ) = det ([ ]−[ ]) = det ([ ])
−3 6 0 λ −3 6−λ
​ ​ ​ ​ ​ ​

And
2−λ 1
det ([ ]) = (2 − λ) ⋅ (6 − λ) − (−3) ⋅ 1 = λ2 − 8λ + 15
−3 6−λ
​ ​

7. Consider the following matrix: 0 / 1 point

3 2
M =[ ]
5 8
​ ​

The covariance matrix related to this matrix is:

Hint: Remember you need to centralize M for each column to first get the matrix denoted in lectures as X , then use the correct
formula. You may want to watch again the lecture on <Add proper lecture>.

2 6
[ ]
6 18
​ ​

0.5 −1.5
[ ]
−1.5 4.5
​ ​

36 46
[ ]
46 68
​ ​

10 −10
[ ]
−10 10
​ ​

Incorrect
This matrix is the result of applying the formula, but computing the mean for each row and not for each column! You may
want to watch again the lecture on PCA - Mathematical Formulation .

8. Consider the following matrix 0.75 / 1 point

3 0
M =[ ]
−2 1
​ ​

Check all the options that represent the eigenvectors of this matrix.
Hint:
The characteristic polynomial for M is given by (3 − λ)(1 − λ).
Remember that for each eigenvalue, if there is a non-zero eigenvector related to it, then there are infinitely many more eigenvectors
related to the same eigenvalue. In other words, if v is an eigenvector for an eigenvalue λ, then kv is also an eigenvector for the same
eigenvalue λ, for any real valued number k .
You may want to watch again the lecture on "Eigenvalues and eigenvectors ".
 1
[ ].
3
​

k
[ ], for any k real.
k
​

This should not be selected

Vectors of such form are not eigenvectors for any eigenvalue of M . Review the lecture on "Eigenvalues and eigenvectors "
to get a better undesrstanding on how to solve such problem.

k
[ ], for any k real.
−k
​

Correct
You first find the eigenvalues for the matrix M . This is given by the characteristic polynomial, and the hint shows us that the
eigenvalues are λ = 3 and λ = 1.
For λ = 3, you must solve
3 0 x x
[ ][ ] = 3[ ].
−2 1 y y
​ ​ ​ ​

3x = 3x
This translates to { . The first equation is always true, the second equation leads to x = −y . So if x = k , then
−2x + y = 3y
​

k
y = −k . Thus any eigenvector is of [ ].
−k
​

0
[ ], for any k real.
k
​

Correct
You first find the eigenvalues for the matrix M . This is given by the characteristic polynomial, and the hint shows us that the
eigenvalues are λ = 3 and λ = 1.
For λ = 1, you must solve
3 0 x x
[ ][ ] = 1[ ].
−2 1 y y
​ ​ ​ ​

3x = x
This translates to { . The first equation implies x = 0 and the second equation leads to y = y , which is always
−2x + y = y
​

0
true, so setting y = k , any eigenvector related to this eigenvalue is of the form ( ).
k
​

9. Suppose you have the following dataset 1 / 1 point

2
Size (m )
No. Bedrooms No. Bathrooms
House 1 ​ 70 2
​ 2 ​ ​ ​

House 2 110 4 2
Which matrix is the X − μ matrix, used in the covariance matrix computation? The matrix X − μ is defined in the lecture
1
Covariance Matrix . Remember that the covariance matrix is defined by Σ = n−1 (X − μ)T (X − μ). ​

70 2 2
X −μ=[ ]
110 4 2
​ ​ ​

−20 −1 0
X −μ=[ ]
20 1 0
​ ​ ​

⎡−20 20⎤
X − μ = ⎢ −1 ​ ​ 1⎥ ​ ​

⎣ 0 0⎦

⎡70 110⎤
X −μ=⎢2 ​ ​ 4 ⎥ ​ ​

⎣2 2 ⎦

Correct
 The matrix that represents the dataset is
70 2 2
M =[ ]
110 4 2
​ ​ ​

By computing the mean for each column, you get that

μx = 90, μy = 3, μz = 2
​ ​ ​

Therefore,
70 − 80 2−3 2−2 −20 −1 0
X −μ=[ ]=[ ]
110 − 80 4−3 2−2 20 1 0
​ ​ ​ ​ ​ ​

10. For the dataset from question 9, what are the eigenvalues of the covariance matrix? 1 / 1 point

λ1 = 802, λ2 = 0, λ3 = 0
​ ​ ​

λ1 = 0, λ2 = 0
​ ​

λ1 = 17027, λ2 = 0, λ3 = 0
​ ​ ​

Correct
−20 −1 0
That is correct. From the previous question, you know that X −μ=[ ], so that the covariance
20 1 0
​ ​

matrix is
1 ⎡800 40 0⎤
C= (X − μT )(X − μ) = ⎢ 40 −2 0⎥
n−1
​ ​ ​ ​ ​ ​

⎣ 0 0 0⎦
Therefore,

⎡800 − λ 40 0 ⎤
det (C − λI ) = ⎢ 40 ​ ​ −2 − λ ​ 0 ⎥ ​ ​

⎣ 0 0 0 − λ⎦
and
det(C − λI) = (800 − λ) ⋅ (2 − λ) ⋅ (0 − λ) + 40 ⋅ 0 ⋅ 0 + 40 ⋅ 0 ⋅ 0−
Solving for det(C − λI) = 0 you get
0 ⋅ (2 − λ) ⋅ 0 − (800 − λ) ⋅ 0 ⋅ 0 − 40 ⋅ 40 ⋅ (0 − λ)
​ ​

the desired eigenvalues.