0% found this document useful (0 votes)
79 views39 pagesChapt 5 Joint ProbabilityDistribution
Uploaded by
Will TedjoCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
79 views39 pagesChapt 5 Joint ProbabilityDistribution
Uploaded by
Will TedjoCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
152.
iStockphoto
Joint Probability Distributions
Air quality monitoring stations are maintained throughout Maricopa County, Arizona and
the Phoenix metropolitan area. Measurements for particulate matter and ozone are meas-
ured hourly. Particulate matter (known as PMO) is a measure (in g/m ) of solid and liq-
uid particles in the air with diameters less than 10 micrometers. Ozone is a colorless gas
with molecules comprised of three oxygen atoms that make it very reactive. Ozone is
formed in a complex reaction from heat, sunlight, and other pollutants, especially volatile
organic compounds. The US. Environmental Protection Agency sets limits for both PM10
and ozone. For example, the limit for ozone is 0.075 ppm. The probability a day in Phoenix
exceeds the limits for PM10 and ozone is important for compliance and remedial actions
with the county and city. But this might be more involved that the product of the probabili-
ties for each pollutant separately. It might be that days with high PM10 measurements also
tend to have ozone values. That is, the measurements might not be independent and it is the
joint relationship between these measurements that becomes important. The study of prob-
ability distributions for more than one random variable is the focus of this chapter and the
air quality data is just one illustration of the ubiquitous need to study variables jointly.
CHAPTER OUTLINE
5-1 TWO OR MORE RANDOM 5-2 COVARIANCE AND
VARIABLES CORRELATION
5.1.1 Joine Probability Distributions
1.2 Marginal Probability Distributions
5-1.3 Conditional Probability
Distributions
COMMON JOINT DISTRIBUTIONS
53.1. Maltinomial Probability Distribution
5.3.2. Bivariate Normal Distribution
5-4 LINEAR FUNCTIONS OF
5-14 Independence RANDOM VARIABLES
5-1.5 Mote Than Two Random 5.5 GENERAL FUNCTIONS OF
Variables
RANDOM VARIABLES 5:1 TWOORMORE RANDOM VARIABLES 153,
LEARNING OBJECTIVES
After careful study of this chapter you should be able to do the following:
|. Use joint probability mass fans
1 38 and joint probability density functions to calculate probabilities
2. Caleulate marginal and conditional probability distributions from joint probability distributions
3. Interpret and calculate covariances and correlations between random variables
4. Use the multinomial distribution to determine probabilities
5. Understand properties of a
probability density function
ivariate normal distribution and be able to draw contour plots for the
6. Calculate means and variances for linear combinations of random variables and calculate proba-
bilities for linear combinations of normally distributed random variables
. isribution of a general function of a random variable
Determine the
In Chapters 3 and 4 we studied probability distributions for a single random variable, However,
it is offen useful to have more than one random variable defined in a random experiment. For
example, in the classification of transmitted and received signals, each signal can be classified
as high, medium, or low quality. We might define the random variable to be the number of
high-quality signals received and the random variable ¥ to be the number of low-quality signals
received. In another example, the continuous random variable X can denote the length of one
dimension of an injection-molded part, and the continuous random variable ¥ might denote the
length of another dimension, We might be interested in probabilities that can be expressed in
terms of both X and ¥. For example, ifthe specifications for X and ¥ are (2.95 to 3.05) and (7.60
to 7.80) millimeters, respectively, we might be interested in the probability that a part satisfies
both specifications; that is, P(2.95 < X-< 3.05 and 7.60 < ¥ < 7.80).
Because the two random variables are measurements from the same part, small disturbances
in the injection-molding process, such as pressure and temperature variations, might be more
likely to generate values for X and ¥ in specific regions of two-dimensional space. For example,
a small pressure increase might generate parts such that both X and ¥ are greater than their
respective targets, and a small pressure decrease might generate parts such that Xand ¥ are both
less than their respective targets. Therefore, based on pressure variations, we expect that the
probability ofa part with X much greater than its target and Y much less than its target is small
In general, if X and Y are two random variables, the probability distribution that defines
their simultaneous behavior is called a joint probability distribution. In this chapter, we
investigate some important properties of these joint distributions
5-1 TWO OR MORE RANDOM VARIABLES
5-1.1 Joint Probability Distributions
For simplicity, we begin by considering random experiments in which only two random vari-
ables are studied. In later sections, we generalize the presentation to the joint probability
distribution of more than two random variables.
EXAMPLE 5-1 Signal Bars
Calls are made to check the airline schedule at your departure In the frst four bits transmitted, let
city. You monitor the number of bars of signal swength on your X denote the number of bars of signal strength
cell phone and the number of times you have to state the name ‘on your cel! phone
of your departure city before the voice system recognizes the
{¥ denote the numberof times you need to state
name your departure city 154 CHAPTERS JOINT PROBABILITY DISTRIBUTIONS
“Fe number of bare of signal stvengib
a 2 2
Figure 5-1 Joint [05 or 005
probability distribution [0.02 on 0.05.
of Nand Yin Example 2/002 0.08 02,
1 roo. 02 025
By specifying the probability of each of the points in Fig. 5-1, range of the random variables (X, ¥) to be the set of points
wwe specify the joint probability distribution of X and ¥. (x) in two-dimensional space for which the probability that
Similarly wo an individual random variable, we define the "= x and ¥ = yi positive
‘Mass Function
IfXand ¥ are discrete random variables, the joint probability distribution of X and Vis a
description of the set of points (x, y) in the range of (X, ¥) along with the probability of each
point, The joint probability distribution of two random variables is sometimes referred to as
the bivariate probability distribution or bivariate distribution of the random variables.
One way to describe the joint probability distribution of two discrete random variables is
through a joint probability mass funetion, Also, P(X = x and ¥ = y) is usually written as
P= Y= 9),
The joint probability mass function of the diserete random variables X and ¥,
denoted as fay(x »), satisfies
fers») = 0
2) LY hole»)
@) fins») = P= 2, Y= y) on
Just asthe probability mass fumetion of a single random variable X is assumed to be zero at all
values outside the range of X, so the joint probability mass function of X and ¥ is assumed to
bee zero at values for which a probability is not specified.
The joint probability distribution of two continuous random variables X and Y can be
specified by providing a method for calculating the probability that X and ¥ assume a value in
any region R of two-dimensional space. Analogous to the probability density function of a
single continuous random variable, a joint probability density function can be defined over
two-dimensional space. The double integral of fyy(x, ») over a region R provides the proba-
bility that (X,Y) assumes a value in R. This integral can be interpreted asthe volume under the
surface fy(x,)) over the region R.
‘A joint probability density function for X and Y is shown in Fig, 5-2. The probability
that (¥,¥) assumes a value in the region R equals the volume of the shaded region in
Fig. 5-2. In this manner, a joint probability density function is used to determine probabil-
ities for X and ¥. fev’
Figure 5-2 Joint probability density function for
random variables X and ¥. Probability that Wf, ¥) is
in the region R is determined by the volume of
5:1 TWOORMORE RANDOM VARIABLES 155
fais»)
3
2% 30
769 5.08
Figure 5-3 Joint probability density fumetion for the lengths
of different dimensions of an injection-molded part.
Sark, 3) over the region
Deasity
Function
A joint probability density function for the continuous random variables X and ¥,
denoted as fyy(x, y), satisfies the following properties:
(D) for(s,y) = 0 for all x, y
Q) | | fire.» ray = 1
(3) For any region R of two-dimensional space,
rene n || frls»aray
(5-2)
‘Typically, fey(x,») is defined over all of two-dimensional space by assuming that
_fiy(%y) = 0 for all points for which fyy(x, y) is not specified.
‘At the start of this chapter, the lengths of different dimensions of an injection-molded part
‘were presented as an example of two random variables. Each length might be modeled by a
normal distribution. However, because the measurements are from the same part, the random
variables are typically not independent. A probability distribution for two normal random vari-
ables that are not independent is important in many applications and itis presented later inthis
chapter. Ifthe specifications for X and Yare 2.95 to 3.05 and 7.60 to 7.80 millimeters, respec-
tively, we might be interested in the probability that a part satisfies both specifications; that is,
P(2.95 < X< 3.05, 7.60 < ¥ < 7.80). Suppose that fyy(,») is shown in Fig. 5-3. The re-
quired probability is the volume of f(x,y) within the specifications. Often a probability such
as this must be determined from a numerical integration.
EXAMPLE 5-2 Server Access Time
Let the random variable denote the time until a computer
server connects to your machiae (in milliseconds), and let Y
denote the time until the server authorizes you as a valid user
(Go milliseconds). Each of these random variables measures
the wait from a common starting time and X< Y. Assume that
the joint probability density function for X and Yis
Fert 3) = 6X 10-%exp(—0,00Lx — 0.0029) fore 2000, °
2000
2000 *
‘The first intogral is
=
6x 108
‘0.002
oe
6x 107
‘002
= 0047s
‘007
°) ca
“The second integral is
wwe | (Si
6X10 Fone
Som |e
6x 08/ et 5
Sear (Sas) 908
‘Therefore,
P(Y > 2000) ~ 0.0475 + 0.0025 ~ 0.08
Alternatively, the probability can be calculated from the mar-
ginal probability distribution of Yas follows. For y > 0,
stm feta
)
We have obtained the marginal probability density funetion
of [Link],
PO > 2000) = 6 x 107 | one —
so (Sr )-
00 a
Also, E(X) and V(X) can be obtained by first calculating the marginal probability disti-
bution of ¥ and then determining EX) and V(X) by the usual method. In Fig. 5-6, the marginal
probability distribution of X is used to obtain the mean as
E(X) = 1(0.2) + 2(0.25) + (0.55) = 2.35
5-1.3 Conditional Probability Distributions
‘When two random variables are defined in a random experiment, knowledge of one can change
the probabilities that we associate with the values of the other. Recall that in Example 5-1, X
denotes the number of bars of service and Y denotes the number of times you need to state 5:1 TWOORMORE RANDOM VARIABLES 159
the name of the departure city. One expects the probability ¥ ~ 1 to be greater at X— 3 bars
than at
such conditional probabil
bar. From the notation for conditional probability in Chapter 2, we can write
s as PC
1X = 3) and P(Y = 114 = 1), Consequently, the ran-
dom variables X and ¥ are expected to be dependent. Knowledge of the value obtained for X
changes the probabilities associated with the values of ¥.
Recall that the definition of conditional probability for events 4 and B is P(BLA) =
PCAN BPA). T
B defined to be
».
EXAMPLE 5-5 Signal Bars
For Example §-1, and ¥ denote the number of bars of signal
strength and times you need to slate your departure city re
ceived, respectively. Then,
PUY = 1 = 3) = POX = 3, = 1/POK = 3)
= fir, Difel®) = 0.25/0.55
454
‘The probability that ¥ = 2 given that X = 3 is
definition can be applied with the event 4 defined to be X = x and event
POP = 21x = 3) = PX = 3, ¥ = 2/P(= 3)
= fia(3, Vf) = 02/055 = 0.364
Additional Conclusion: Further work shows that P(Y = 3]X'=
3) = 0.091 and P(Y = [X= 3) = 0.091, Note that P(Y = 1]X
') + P(Y = 2|X = 3) > P(Y = 3[X = 3) + P(Y = 41K
3) = |. This set of probabilities defines the conditional proba-
bility distribution of ¥ given that X= 3.
Example 5-5 illustrates that the conditional probabilities for ¥ given that X = x can be
thought of as a new probability distribution. The following definition applies these concepts to
continuous random variables.
Conditional
Probability
Density
Fun
Fic)
Given continuous random variables X and Y with joint probability density function
Fer(%,y), the conditional probability density function of ¥ given X = xis
fer)
for fx(x) > 0 G4)
Sele)
values of ¥ given that X = x.
‘The conditional probability density function provides the conditional probabilities for the
@) fuoy=0
@® [ioe
Because the conditional probability density function fy,(y) is a probability density
function for all y in R,, the following properties are satisfied:
6) pyresix=.)= [ssb)ay for any set B in the range of Y
(5-5) 160 CHAPTER JOINT PROBABILITY DISTRIBUTIONS
Figure 5-8 The y
‘conditional probability
density funetion for ¥,
igen that x = 1500, is
nonzero over the solid
line.
1500
o
‘01500 z
It is important to state the region in which a joint, marginal, or conditional probability
density function is not zero. The following example illustrates this.
EXAMPLE 5-6 Conditional Probability
For the random variables that denote times in Example 5-2,
determine the conditional probability density fuaction for Y
given that X =x.
First the marginal density function of x ig determined.
Forx> 0,
fel)
[ox irteamesany
(
for
= 6x 10r6er8008
= 0.0036 800%
‘This is an exponential distribution with 4 = 0,003. Now, for
0. < xand x < ythe conditional probability density function is
x>0
6% 10~fe80012~ 0.025
0.003
for 0 2000 |X = 1500)
‘The conditional probability density function is integrated as
follows:
Py > 2000pt = 1500) = | frysls)
[ ecoreesm-n gy
a)
~ vane (1) - 038
=o
- ox2e(
Properties of random variables can be extended to @
conditional probability distribution of ¥ given X = x. The
‘usual formulas for mean and variance can be applied to a
conditional probability density or mass function,
“== number of bars of signal strength
Figure 5-9 1 2 3
Conditional probabil- 0760 400 [ooa
ity distributions of ¥ 3 o200 ‘a400 [ooar
given X=, fyil») 20.100 0.20 [064
in Example 5-7 10.050 080 [os Conditional
Mean and
Variance
5:1 TWOORMORE RANDOM VARIABLES 161
The conditional mean of ¥ given X = x, denoted as B(Y |x) oF tytn is
6-6)
eerie) = [ino
and the conditional variance of ¥ given X = x, denoted as V(¥|x) or 6},,, is
nyin)= [= wat find) = [Anode
EXAMPLE 5-8 Conditional Mean and Variance
For the random variables that denote times in Example 5-2,
determine the conditional mean for Y given thatx = 1500,
The conditional probability density funetion for ¥ was
determined in Example 5-6, Because fy») i8 nonzeeo for
> 1500,
B(Y|X = 1500) = | (0.002 8209-900 dy
EXAMPLE 5.9
For the discrete random variables in Example 5-1, the condi-
tional mean of ¥ given X'~ | is obtained from the conditional
distibution in Fig 5-9
BN) = way
1(0.08) + 2(0.1) + 3(0.1) + 4(0.78)
=355
5-1.4 Independence
Integrate by parts as follows:
0.002
[roma
1500
02"
1500
D002
ae (2000)
(0.002)(0.002)
With the constant 0,002 reapplied,
Practical Interpretation: Ifthe connect time is 1500 ms then
the expected time to be authorized is 2000 ms,
‘The conditional mean is interpreted asthe expected aumber of
times the city name is stated given that one bar of signal is
present, The conditional variance of F given X = is
YI) = 0 — 3.55) 0.05 + (2 — 355)%01
+ (3 — 3550.1 + (4 — 355)°0.75 = 0.748
In some random experiments, knowledge of the values of X does not change any of the prob-
abilities associated with the values for ¥.
EXAMPLE 5-10 Independent Random Variables
In plastic molding operation, each part is classified as to
‘whether it conforms to color and length specifications, Define
the random variable X'and Yas
1
efi
mf
if the part conforms to color specifications
otherwise
if the part conforms to length specifications
otherwise
Assume the joint probability distribution of Xand Y's,
defined by fy, in Fi. 5-10(a) The margingl probab
ity distributions of ¥ and Yare also shown in Fig. 5-10(a).
Note tat fer(t, 3) = fee) fr(0). The conditional probabi
ity mass function fy(9) is shown in Fig, 5-10(6). Notice
that for any x, fy1.(9) ~ fr(0). Thetis, knowledge of whether
‘or not the part meets color specifications does not change the
probability that it meets length specifications. 162 CHAPTER JOINT PROBABILITY DISTRIBUTIONS
Figure 5-10 (a) Joint y y
and marginal probabil. fy6 =
ity distributions of ¥ ggg 1 42998, 99702 11998088
Sovinbwnnle
S10.) Conia oma pore oor 002
rotbityydsibuien 992 2 ee °F pe
werenentee oy ons
Basis
« »
By analogy with independent events, we define two random variables to be independent
whenever fey(t, y) = fats) f(y) for all x and y. Notice that independence implies that
Sarbss 9) = fide) frly) for all x and y. If we find one pair of x and y in which the equality fails,
‘Xand ¥ are not independent, If two random variables are independent, then for fy(x) > 0,
fls.9) _ fs) 0)
fil) = BONO 50
OR aay)
{With sma ealeultions, the fllowing equivalent statements canbe shown
Independence
For random variables X and Y, if any one of the following properties is true, the
others are also true, and X and Y are independent.
(CD) fers») = fi) fr) forall xand y
2) fye(y) = fr) for all x and y with f(x) > 0
3) fay) — fel) forall xand y with fy(y) > 0
(4) POLE A, Y © B) = PIXE A)P(Y € B) for any sets A and B in the range
of Xand ¥, respectively. 7)
Rectangular Range for (X; Y)
If the set of points in two-dimensional space that receive positive probability under
Ferlx, 9) does not form a rectangle, X and ¥ are not independent because knowledge of X
can restrict the range of values of ¥ that receive positive probability. If the set of points in
two-dimensional space that receives positive probability under fer(x, ») forms a rectangle,
independence is possible but not demonstrated, One of the conditions in Equation 5-6 must
still be verified.
‘The variables in Example 5-2 are not independent. This can be quickly determined be-
cause the range of (X, ¥) shown in Fig. 5-4 is not rectangular. Consequently, knowledge of X
changes the interval of values for ¥ with non-zero probability.
EXAMPLE 5-11 Independent Random Variables
se iatiy hase of hand? oie oer i) [axiotessinene gy
10™exp(—0.001x=0,002y) for x= 0 and y = 0, Show that X }
ease eee ener tour tom
‘The marginal probability density function of X is = O.001e°8F fore > 0 ‘The marginal probability density function of Yis
fy) = [2 1o-ten—am de
= 0.002679 fury > 0
1085 Y) = fal) 9) forall x and y, and Xand Y
are independent,
Often, based on knowledge of the system under study, random variabl
5:1 TWOORMORE RANDOM VARIABLES 163
‘To determine the probability requested, property (4) of
Equation 5-7 and the fact that each random variable has an
‘exponential distribution can be applied
P(X > 1000, ¥ < 1000) = P(X > 1000}P(¥ < 1000)
NL =e?) = 0.318
s are assumed to
be independent, Then, probabilities involving both variables can be determined from the mar-
ginal probability distributions. For example, the time to complete a computer search should be
independent of an adult’ height.
EXAMPLE 5-12 Machined Dimensions
Let the random variables X and ¥ denote the lengths of two di
rensions of a machined par, respectively. Assume that Xand
Yare independent random variables, and further assume that
the distribution ofX is normal with meas 10.5 millimeters and
variance 0,0025 (millimeter)? and that the distribution of
Yis normal with mean 3.2 millimeters and variance 0,036
(cillimeter, Determine the probability that 10.4 , X3,
the multiple integral of fx, x, (t1. 2.
»%,) assume a value in a region R by
+ %,) over the region R.
Joint
Probability
Dens
Fun
A joint probability density function for the continuous random variables X\, X3,
Hy, oo Xp denoted 88 fix. (eX
(D Figay 40800100 %) =O
@ {J | sa 4g (ip Mayo Hp) dey dy dtp
(3). For any region B of p-dimensional space,
PUM Xa.» Xp) € B] =
1%), satisfies the following properties:
[foe [peas scons es es dip
. 8) 164
‘Typically, fy, x,...x, (15 ¥25
Fist
EXAMPLE 5-14 Component Lifetimes
In an electronic assembly, let the random variables
Xi XaXs, Xa denote the lifetimes of four components in
hours. Suppose that the joint probability density function of
these variables is
sess ln % 85.84)
ey Tone = ts 0002115, OOS
form = 0.x; = 0,45 = Ox, = 0
‘What is the probability thatthe device operates for more than
1000 hours without any failures? The requested probability is
CHAPTER S JOINT PROBABILITY DISTRIBUTIONS
+p) = 0 forall points for which fix, 015%»
+ %,) 8 defined overall ofp-dimensional space by assuming that
1%) isnot specified
PUY > 1000, Xz > 1000, X5 > 1000, X, > 1000), which
‘equi the multiple integral of fxn) Over the
region x, > 1000, x > 1000, % > 1000, 4 > 1000. The joint
probebilty density fmetion can be written as « product of
‘exponential functions, and each integral isthe simple integral
cof an exponential function. Therefore,
P(X, > 1000, %3 > 1000, %5 > 1000, X4 > 1000)
153 = 0.00055
Suppose that the joint probability density function of several continuous random
variables is a constant, say, c over a region R (and zero elsewhere). In this special ease,
[ff as alert de dy
X (volume of region R)
by property (2) of Equation 5-8, Therefore, (R). Furthermore, by property (3) of Equation 5-8,
PUK Xo) © B]
volume (BR)
volume (R)
[J [on ats a ay 0 some 8
EXAMPLE 5-15. Probabil
Suppose the joint probability density function of the con-
tinuous random variables X and Y is constant over the re- ra
gion x +)7=4, Determine the
eee
Function
‘When the joint probability density function is constant, the probability that the random vari-
ables assume a value in the region B is just the ratio of the volume of the region B MR to the
volume of the region R for which the probability is positive.
ty as a Ratio of Volumes
‘The region that receives positive probability is citcle of
vs 2. Therefor, the area ofthis region is 4. The area of
the region x? + y° 1 is . Consequently, the requested
probability is «/(4n) = 1/4,
probability that
Ifthe joint probability density function of continuous random variables Xj, Xp... Xp
48 fizx4...% (1s 42 ++ Xp), the marginal probability density function of X, is
(5-9)
ded = [J | fan lay Xap vee 9 Mp) By Ag oo iin Bing oe tp
‘where the integral is over all points in the range of X;, Xo, ... , Xp for which X; Mean and
Variance from
Joint Distribution
EXAMPLE 5-16
5:1 TWOORMORE RANDOM VARIABLES 165
As for two random variables, @ probability involving only one random variable, say, for
example, Pla 0.
The concept of independence can be extended to multiple random variables.
Independence
Random variables 4, X,
Sit 92%) =
AX, are independent if and only if
fer)
Filip) forall x1, 5... .%
6-12)
Similar to the result for only two random variables, independence implies that Equation 5-12
holds for all x, x
PU, © Ay Xp © Ay,
for any regions Ay, Aa... Ay
EXAMPLE 5-17
In Chapter 3, we showed that a negative binomial random vari=
able with parameters p and r can be represented as a sum ofr,
‘geometric random variables 1, Xa,-.-.,. Bach geometric
random variable represents the additional trials required 10
EXAMPLE 5-18 Layer Thickness
Suppose X,, 13, and X; represent the thickness in micrometers
of a substrate, an active layer, and a coating layer of a chemical
product, Assume that X;,.X, and X; are independent and not-
‘mally distributed with jy = 10000, jy = 1000, yy = 80,
‘0 = 250, a2 = 20, and 0 = 4, respectively. The specifica
tions for the thickness of the substrate, active layer, and cout-
ing layer are 9200 ,
1X, are
» X; are independent,
Xp E Ap) = P(X, E Ai)P(X2 E Az)... PX E Ap)
_X,, respectively.
dobtain the next success. Because the trials in a binomial
experiment are independent, ,%3,...,%, are independent
random variables,
‘The requested probability is P(9200 42)
), BU), 1X), and HCY).
(£) Marginal probability distribution of the random vari=
able x
(g) Conditional probability distribution of ¥ given that X= 1.5,
(h) Conditional probability distribution of X given that Y = 2
@ BUX = 1.5)
() Ate X and ¥ independent?
5-2, Determine the value of ¢ that makes the funtion
‘fle,y) = eG + y) a joint probability mass fumetion over the
nine points with x = 1,2,3 andy = 1,2,3.
Determine the following:
@ P= 1 <4) &) P= 1)
(© PY =2) (@ PUW<2,¥<2)
(©) EQ), E(Y), V0), and 17)
(£) Marginal probability distribution of the random variable X'
(g) Conditional probability distribution of ¥ given that X = 1
(1) Conditional probability distribution of X given that
@ AYIX= 1)
(j) Ate Xand ¥ independe:
5-3, Show that the following funetion satislies the proper-
ties ofa joint probability mass function.
x » far)
=I 2 1/8
05 1 Ws
0s 1 2
1 2 18
Determine the following
(a) P< 05, < 13)
() P= 03)
© AY<15)
(@) PU> 025, ¥< 45)
(@) EX), EY), 1), and 1)
(£) Marginal probability
ableX
(2) Conditional probability distbution of Y given that
x=1
(8) Conditional probability distribution of X given that
yor
@ ElXly = 1)
() Are Xand Findependent?
5-4. Four electron printers are selected from a large lot
of damaged printers. Each printer i inspected and classified
48 conning either a major ora minor defect. Let the random
variables X and Y denote the numberof printers with major
and minor defects, respectively. Determine the range of the
joint probability distibution of X and ¥
5.5. Inthe transmission of digital information, the probability
thatabit has high, modsrate, and low distortion is 0.01, 0.04, and
0.95, respectively: Suppose tha three bits are transmitted and
that the amount of distortion of each bit is assumed to be
independent. Let X and ¥ denote the number of bits with high
and moderate distortion out ofthe thee, respectively. Determine:
bution of the random varie
(®) fits) 0) fa)
(©) EX) @ flo)
(©) BOX = 1) (0) Are X and ¥ independent?
5-6. A small-business Web site contains 100 pages and
(60%, 30%, and 10% of the pages contain low, moderate, and
high graphic content, respectively. A sample of four pages is
selected without replacement, and X and ¥ denote the number
of pages with moderate and high graphics output in the
sample. Determine:
(@) firs») ) AE)
(©) EX) @) frs 0)
(©) BY x= 3) © Mlr=3)
(g) Ate Xand Vindependent?
5-7. A manufacturing company employs two devices to
inspect ouput for quality control purposes. The first device
is able to accurately detect 99.3% of the defective items it
receives, whereas the second is able to do so in 99.7% of the
ceases, Assume that four defective items are produced and
sent out for inspection. Let X and ¥ denote the number of
items that will be identified as defective by inspecting 168 CHAPTERS JOINT PROBABILITY DISTRIBUTIONS
devices 1 and 2, respectively. Assume the devices are inde-
pendent. Determine:
©) firls2) ©) Fels)
(© EX) @ fr2)
(©) BUX = 2) () X= 2)
(g) Are Xand Yindependent?
5-8. Suppose the random variables X, ¥, and Z have the
following joint probability distribution,
L692)
0.05)
0.10
os
0.20
020
os
0.10
0.05
peeenenak
Determine the following
(@ PO 2) (@) PUK= y= 2)
© Pz<13) (@) Par=1 or 2=2)
© EW) () Pa 1|Y=1)
@) P= LY =112=2) dy P= Y= 1,
(i) Conditional probability distribution of X given that
andZ
5-9. An engineering statistics class has 40 students and
60% are electrical engineering majors, 10% are industrial
‘engineering majors, and 30% are mechanical engineer
majors. A sample of four students i selected randomly, with=
‘ul replacement, fora project team. Let X and ¥ denote the
‘numberof industrial enginccring and mechanical cngincering
‘majors respectively, Determine the following
(©) Sayles) ) fel)
© EX) @ fis)
fe AUX =3) (0) WIX= 3)
(g) Are Xand Yindependent?
“Average Frequencies and Operations in TPC-C
5-10. An article in the Journal of Database Manage-
ment (“Experimental Study of a Self-Tuning Algorithm for
DBMS Butler Pools” (2005, Vol. 16, pp. 1-20)) provided
the workload used in the TPC-C’ OLTP (Transaction
Processing Performance Council's Version C On-Line
‘Transaction Processing) benchmark, which simulates a typ-
ical order entry application. Sce the table below. The fre-
quency of each type of transaction (in the second column)
can be used asthe percentage of each type of transaction
Let X and Y denote the average number of selects and
updates operations, respectively, required for each type
transaction, Determine the following:
@) A= 5)
©) ER)
(6) Conditional probability mass function of X given ¥ = 0
(@) POX < 6|Y = 0)
© EalY= 0)
Sel1. For the Transaction Processing Performance
Council's benchmark in Exereise 5-10, let, Y,andZ denote
the average number of selects, updates, and inserts opera-
tions required for each type of transaction, respectively
Caleulate the following:
() firals.9.2)
(©) Conditional probability mass Function for Yand Y given
a
(© PR <6¥ <6!
(@) EXIT =0,
5-12. In the transmission of digital information, the
probability that abit has high, moderate, or low distortion is
4.0, 0.04 and 0.95, respectively. Suppose that three bit are
transmitted and thatthe amount of distortion ofeach bit is as-
sumed to be independent. LetX and Y denote the number of
bits with high and moderate distortion out ofthe thee trans-
ited respectively. Determine the following:
(a) The probability that wo bits have high distortion and one
has moderate distortion
(©) The probability that all hee bits have low distortion
(6) The probability distribution, mean, and vatiance of X
(d) The conditional probability distribution, conditional mean,
and conditional variance of¥ given that Y= 2
Non-Unique
‘Transaction Frequency Selects _Updates_Inserts Deletes Selects __Joins
New Order 3 a "1 2 0 0 0
Payment “4 42 3 1 0 06 0
(Order Status 4 4 0 o 0 06 0
Delivery 5 130 10 oo 0 o
Stock Level 4 0 0 0 0 0 1 5-13. Determine the value of © such that the funetion|
Fos, )) = ony for 0-< x<3 and 0-18,1<¥<25)
(@) 5X) (6) P< 0,9 <4)
{g) Marginal probability distribution ofthe random variable X
{(h) Conditional probability distribution of Y given that X'= 1.5
(i) EOIX) = 15) G) POY < 21x = 18)
‘(k) Conditional probability distribution of X given that Y 5 2
5-14. Determine the value of that makes the function
Js, 9) = ele + y) a joint probability density function over the
range 1) (@) PU <2, <2)
© Fe) MX)
{g) Marginal probability distribution of X
{h) Conditional probability distribution of ¥ given shat X = 1
@ BYIX=1)
() POY> 2\x= 1)
‘) Conditional probability distribution of X given that Y 5 2
5-15. Determine the value of ¢ that makes the function
fls.») = ea joint probability density function over the range
Oex<3and0 1) (@) P< 2,7 = 2)
fe) BX) () EO)
(g) Marginal probability distibution of X
{(h) Conditional probability distribution of ¥ given X = 1
@ EWx=1)
() Pr> 21X= 1)
(6) Conditional probability distribution of X given YS 2
5-16. Determine the value of that makes the funtion
Js.) ~ ce" a join probability density funtion over the
range 0 < xand 0 3) (@ P< 2, =2)
@ BO (9 £0)
{(@) Marginal probability distribution of X
{(h) Conditional probability distribution of ¥ given X = 1
x= 1)
{G) Conditional probability distribution of X given ¥ = 2
Determine the value of ¢ that makes the fimetion
F(2,3) = ce" a joint probability density function over
the range 0 < x and.x < y.
Determine the following:
(@ PW <1, 7 <2)
() PSX <2)
5:1 TWOORMORE RANDOM VARIABLES 169
© RY>3) (@ P< 2,7<2)
© FX) © £0)
{(g) Marginal probability distribution of X
(h) Conditional probability distribution of ¥ given X= 1
(@ BYIX=1)
@ PY <2iX= 1)
(Ge) Conditional probability distribution of X given ¥ = 2
5-18. The conditional probability distribution of Y given
X =x is f(y) =xe7” for y > 0, and the marginal
probability distribution of Xs a continuous uniform dstibu-
tion over Oo 10.
(4) Graph fpe(9) = xe7™ for y > 0 for several values of x:
Determine:
(b) POY <21X = 2) (©) BYIX = 2)
@ BUIX=%) © farts)
8) fr)
5-19. ‘Two methods of measuring surface smoothness are
ised to evaluate a paper product, The measurements are
recorded as deviations from the nominal surface smoothness
in coded units. The joint probability distribution of the
two measurements is a uniform distribution over the 1e-
gion 0 < x= 4,0 0,
peO2>Oands+y+2 O-and oy > 0, if the covariance between X and Y is positive, negative, or zero,
the correlation between X and ¥ is positive, negative, or zero, respectively. The following
result can be shown,
For any two random variables X°and ¥,
-1,
(3) The trials are independent.
‘The random variables XX,
, class k, respectively, have a multinomial distribution and the
{joint probability mass function is
Ay =u)
+ xq = mand p; + pp +
4 class k
+ Pr Tespectively.
» X that denote the number of trials that result in
n!
Tema PPP E18)
sila! a
+m=1
‘The multinomial distribution is considered a multivariable extension of the binomial distribution.
EXAMPLE 5-25 Digital Channel
In Example 5-9, lel the random variables Xj, XX, and Xi de~
note the number of bits that are £, G, F, and P, respectively, in
4 transmission of 20 bits. The probability that 12 of the bits
received are E, 6 are G, 2 are F, and 0 are Pis
PO 6X5 = 2%, = 0)
0,670,3°0,08%0,02" = 0.0358
20!
12612101 53 COMMON JOINT DISTRIBUTIONS — 177
Each trial in a multinomial random experiment can be regarded as either generating or not
generating a result in class i, for cach i = 1, 2,
. k Because the random variable X; is the
umber of trials that result in class i, X; has a binomial distribution.
Mean and
Variance | 1fX, Xp,
tion of X, is binomial with
EX) = mp,
Xi have # multinomial distribution, the marginal probability distribu-
and V(X) = npi( — pi) 19)
EXAMPLE 5-26 Marginal Probability Distributions
In Example 5-25 the marginal probability distribution of Xs is
binomial with n = 20 and p = 0.3. Furthermore, the joint mat-
sina probability distribution of X; and X, is found as follows.
‘The P(X, = x, X; = x) is the probability that exactly x tials re-
sult in G and that x result nF. The remaining w — x, ~ xy tials
‘must result in either F or P, Consequently, we can consider each
tral in the experiment to result in one of three classes, (G}, {F},
‘or {E, P}, with probabilities 0,3, .08, and 0.6 + 0.02 = 0.62,
respectively. With these new classes, we ean consider the trials
5-3.2 Bivariate Normal Distribution
to comprise anew multinomial experiment, Therefore,
figsl2%5) = PO
ay =)
= SE (0 O.08)M (0.
‘The joint probability distribution of other sets of variables can
be found similarly
An extension of a normal distribution to two random variables is an important bivariate prob-
ability distribution,
EXAMPLE 5-27 Bivariate Normal Distribution
At the start of this chapter, the length of different dimensions
‘of an injection-molded part was presented as an example of
‘wo random variables. Each length might be modeled by a not-
mal distribution, However, bectuse the measurements are
from the same part, the random variables are typically not
independent. A probability distribution for two normal ran-
Bivariate Normal
Probability
Density Function
Bros Yi Ox Ty ys Heys P) =
dom variables that are not independent is important in many
applications. As stated atthe start ofthe chapter, ithe speciti-
cations for X and ¥ are 2.95 to 3.05 and 7.60 to 7.80 mille.
ters, respectively, we might be interested in the probability that
4 part satisfies both specifications; that is, P2.98 <.X <
3.05, 7.60 < ¥ < 7.80),
The probability density function of a bivariate normal distribution is
2mayoy"
_ Pole = wx)y = 4)
for— 0,o)>0, 0 < my <=,
2 0
(p <0), the major axis of each ellipse has positive (negative) slope, respectively. If p ~ 0, the
major axis ofthe ellipse is aligned with either the x or y coordinate axis.
EXAMPLE 5-28
‘The joint probability density fmetion fy(,y)= This probability density function is illustrated in Fig. 5-18.
1 osye =p) Notice that the contour plot consists of concentric circles
VE is a special case of a bivariate normal —ghout the origin
distribution with oy = Loy = 1, tye =
ty = 0, and p = 0.
‘The following results can be shown for a bivariate normal distribution. The details are left
as an exercise.
‘Marginal
Distributions of | [f¥and ¥ have a bivariate normal distribution with joint probability density fey(x,»s
Bivariate Normal | gy, gy, ys Hyp), the marginal probability distributions of X and ¥ ate normal
Random Variables | with means jy and pryand standard deviations oy and oy, respectively. (5-21)
Figure 5-19 illustrates that the marginal probability distributions of X and Y are normal.
fortes)
° ° x
Figure 5-18 Bivariate normal probability density Figure 5-19 Marginal probability
function with oy = 1,0) = 1,p=0, 44 = 0,and density funetions ofa bivariate
by=0. ‘normal distribution, Conditional
Distribution of
Bivariate Normal
Random Variables
Correlation of
Bivariate
Random Vi
For Bivariate
Normal Random
Variables Zero
Correlation Implies
Independence
53, COMMON JOINT DISTRIBUTIONS 179.
If X and ¥ have a bivariate normal distribution with joint probability density
Fay ¥s x Fy, hx bey P)> the conditional probability distribution of Y given
X = xis normal with mean
oy
bry = By — Bx Pay
and variance
He = O71 — p*)
Furthermore, as the notation suggests, p represents the correlation between X and Y. The
following result is left as an exercise.
1fX and Yhave a bivariate normal distribution with joint probability density function
Fig ¥i Oy Fy yo hy 6) the correlation between X and Yis p. (5-22)
‘The contour plots in Fig. 5-17 illustrate that as p moves from zero (left graph) to 0.9 (right
graph), the ellipses narrow around the major axis. The probability is more concentrated about
a line in the (x, ») plane and graphically displays greater correlation between the variables. If
6 = —L or +1, all the probability is concentrated on a line in the (x, y) plane. That is, the
probability that X and Y assume a value that is not on the line is zero. In this ease, the bivari-
ate normal probability density is not defined.
In general, zero correlation does not imply independence, But in the special case that X
and Y have a bivariate normal distribution, if p = 0, Xand Y are independent, The details are
left as an exercise.
IfX and Yhave a bivariate normal distribution with
0, Xand Yare independent.
(5-23)
‘An important use of the bivariate normal distribution is to calculate probabilities involving
two correlated normal random variables.
EXAMPLE 5-29 Injection-Molded Part
Suppose that the X’ and ¥ dimensions of an injection-molded
part have a bivariate normal distribution with oy = 0.04,
oy = 0.08, jy = 3.00, wy = 7.70, and p = 0.8. Then, the
probability that a part satisfies both specifications is
P(2.95 < X-< 3.05, 7.60 < ¥< 7.80)
‘This probability can be obtained by integrating fiy(x, 93 0,
thx Hy ) over the region 2.95 denote the length and centimeters and the variance of Yis
‘width, respectively, of a manufactured part. Assume that is
normal with E(X;) = 2 centimeters and standard deviation VO) = 4 x01 14 x 024 = 02
0.1 centimeter, and that is normal with BUYS) = S centime- Noy,
ters and standard deviation 0.2 centimeter. Also, assume that
and X, are independent. Determine the probability that the PV > 145) = PL(Y — py)/ay > (145 — 14)/-VO2)
perimeter exceeds 14.5 centimeters = PZ > 1.12) = 0.13
‘Then, ¥ = 2X; + 2X; is a normal random variable that
represents the perimeter of the part, We oblain E(Y) = 14 184 CHAPTERS JOINT PROBABILITY DISTRIBUTIONS
EXAMPLE 5-33 Beverage Volume
Soft-drink cans are filled by an automated filling machine. The
‘mean fill volume is 12.1 uid ounces, and the standard deviae
tion is 0.1 fluid ounce. Assume tha the fil volumes ofthe cans
are independent, normal random variables. What isthe proba~
bility that the average volume of 10 cans selected from this
proces is less than 12 id ounces?
Let. Xi, Xiy--- sp denote the fill volumes ofthe 10 eas.
‘The average fill volume (denoted as ) is a normal random
variable with
EXERCISES FOR SECTION 5-4
or
and VQ) = 22 0.001
yet ¥
BR) = 12.1 =
Consequently,
12-121
p< 2) =| Sat
Vanor
= PZ < 3.16) = 0.00079
5-54. end Yare independent, normal random variables with
EU) = 0,12) = 4, £0) = 10,and WY) = 9.
Determine the following
(@) EOX +3Y)
(©) POX + 3¥< 30) (d) PAX + 3¥ < 40)
5-55. X and ¥ are independent, normal random variables
with £0) = 2, UK) = 5, BUY) = 6,and FOP) = 8
Determine the following:
(@) EX + 2Y) (0) WGx + 2¥)
(©) PGX + 2¥-< 18) (d) PGX + 2¥-<28)
5-56. Suppose that the random variable X represents the
length of a punched part in centimeters. Let ¥ be the length
ofthe part in millimeters. I1E(X) = Sand V(X) = 0.25, what
are the mean and variance of ¥?
5-57. A plastic easing for a magnetic disk is composed of
two halves. The thickness of each half is normally distributed
With a mean of 2 millimeters and a standerd deviation of
0. millimeter and the halves are independent
(@) Determine the mean and standard deviation of the total
thickness ofthe two halves.
(b) What is the probability thatthe total thickness exceeds
43 millimeters?
5-58. Making handcrafted pottery generally takes two
major steps: wheel throwing and firing. The time of wheel
throwing and the time of fring are normally distributed
random variables with means of 40 min and 60 min and stan-
dard deviations of 2 min and 3 min, respectively.
{(@) What i the probability that a piece of pottery wil be fin-
ished within 95 min?
(@) Whats the probability that t wil take longer than 110 min?
5.59,
eral different layers of ink are deposited onto a plastic sub~
site, The thickness ofthese layers is eitical if specifications
regarding the final color and intensity of light are to be met,
Let Yand ¥ denote the thickness of two different layers of ink.
Itis known that X is normally distributed with a mean of 0.1
rillimeter and a standard deviation of 0.00031 millimeter and
Y'is also normally distributed with a mean of 0.23 millimeter
and a standard deviation of 0.00017 millimeter. Assume that
these variables are independent
() Vex + ay)
In the manufacture of electroluminescent lamps, sev=
(@ Ifa particular lamp is made up of these two inks only,
What isthe probability that the total in
than 0.2337 millimeter?
() A lamp witha total ink thickness exceeding 0.2405 mil-
limeter lacks the uniformity of color demanded by the
customer. Find the probability that a randomly selected
lamp fails to meet eustomer specifications
5-60. The width ofa casing for a door is normally disteib-
‘uted with # mean of 24 inches and a standard deviation of
1/8 inch, The width of a door is normally distributed with «
‘mean of 23-7/8 inches and a standard deviation of 1/16 inch.
‘Assume independence,
(@) Determine the mean and standard deviation of the differ-
cence between the width ofthe casing and the width ofthe
door,
(b) What is the probability thatthe width ofthe cas
the width of the door exceeds 1/4 inch?
(©) What is the probability that the door does not Gt in the
casing?
561. An article in Knee Surgery Sports Traumatology
Arthroscopy (“Effect of Provider Volume on Resource
Utilization for Surgical Procedures” (2005, Vol. 13, pp.
273-279)] showed a mean time of 129 minutes and a standard
deviation of 14 minutes for ACL reconstruction surgery for
high-volume hospitals (with more than 300 such surgeries per
year) If high-volume hospital needs to schedule 10 surger-
ies, what are tae mean and variance ofthe total time to com-
plete these surgeries? Assume the times of the surgeries are in-
dependent and normally distributed
5-62. Softdrink cans are filled by an automated Gilling
‘machine and the standard deviation is 0.5 Guid ounce. Assume
‘that the fill volumes of the cans are independent, normal
random variables
(a) What is the standard devi
of 100 cans?
(&) Ifthe mean fill volume is 12.1 ounces, what is the proba-
bility thatthe average fill volume ofthe 100 cans is below
12-fhuid ounces?
(©) What should the mean fill volume equal so that the proba-
bility thatthe average of 100 cans is below 12 fluid ounces
is 0.0087
ickness is less
n of the average fill volume 55
(@) Ieee meas fil volume is 12.1 ud ounces, what should
te standard deviation of fill volume equal so that the
probability thatthe average of 100 cans is below 12 fuid
ounces is 0.0052
(©) Determine the numberof cans that need fo be measured
Such thatthe probability thatthe average fill volume is
less than 12 Buid ounces is 0.01
5.63. The photoresist thickness in semiconductor mamfac-
‘uring has a mean of 10 micrometers anda standard deviation of
1 micrometer. Assume tat the thickness is normally distributed
and tat the thicknesses of different wafers ae independent
{@) Determine the probability thatthe average thickness of 10
‘wafers either greater than 11 of Tes than 9 micrometer.
(b) Determine the sumber of wafers that need tobe measured
such thatthe probability that the average thickness ex-
ceeds 11 micrometers is 0.01
(©) Ifthe mean thickness is 10 micrometers, what should the
standard deviation of thikness equal so thatthe probability
that the average of 10 wafers is either greater than 11
‘or less than 9 micrometers is 0.001?
5-64. Assume thatthe weights of individuals are indepen-
dent and normally distributed with a mean of 160 pounds anda
standard devition of 30 pounds. Suppose that 25 people
squeeze into an elevator that i designed to hold 4300 pounds.
{@) Whatisthe probability thatthe load (total weight) exceeds
the design limit?
(b) What esign li
ability 0.00012
5-65. Weighs of parts are normally distributed with vari
©, Measurement eror is normally distributed with mean zero
and variance 0.50 independent ofthe part weighs and adds to
te part weight. Upper and lower specifications are centered at
3er about the process mean
tis exceeded by 25 occupants with prob-
55
GENERAL FUNCTIONS OF RANDOM VARIABLES 185
(2) Without measurement error, what is the probability that 2
part exceeds the specifications?
(&) With measurement error, what is the probability thats part
is measured as beyond specifications? Does this imply itis
truly beyond specifications?
(©) What is the probability that a partis measured beyond
specifications ifthe true weight of the partis one o below
the upper specification limit?
5-66. A U-shaped component isto be formed from the three
parts 4, B, and C, The picture is shown in Fig. 5-20, The length
of 4 is normaly distributed with a mean of 10 millimeters and
standard deviation of 0.1 millimeter. The thickness of parts B
and Cis normally distributed with a mean of 2 millimeters and
standard deviation of 0.05 millimeter. Assume all dimensions
are independent,
“2 |- “icf
5 ©
a
A 4
Figure 5-20 Figure for the
U-shaped component
(a) Determine the mean and standard deviation of the length
of the gap D.
(b) What is te probability that the gap D is less than 5.9 mil-
limeters?
GENERAL FUNCTIONS OF RANDOM VARIABLES
In many situations in statisties, itis necessary to derive the probability distribution of a func-
tion of one or more random variables. In this section, we present some results that are helpful
in solving this problem,
Suppose that X is a diserete random variable with probability distribution f(x). Let Y=
(X) be a function of X that defines a one-to-one transformation between the values of X and
Y, and we wish to find the probability distribution of ¥. By a one-to-one transformation, we
‘mean that each value x is related to one and only one value of y ~ f(x) and that each value of
1y is related to one and only one value of x, say, x = u(y), where u() is found by solving y =
‘h(x) for x in terms of y.
Now, the random variable ¥ takes on the value y when X takes on the value u().
‘Therefore, the probability distribution of Y is
Shy) = PY
‘We may state this result as follows.
= y) = PIX = wly)] = Salut) 186 CHAPTERS JOINT PROBABILITY DISTRIBUTIONS
General
Function of | Suppose that X is a diserete random variable with probability distribution f(x). Let
Discrete | y = (X) define a one-to-one transformation between the values of X and ¥ so that
Random Variable | the equation y = h(x) can be solved uniquely for x in terms of y. Let this solution be
x = u(y). Then the probability mass function of the random variable Y is
Sy) = Hu) (5-30)
EXAMPLE 5-34 Function of a Discrete Random Variable
Let X be a geometric random variable with probability disti- ‘Since X= 0, the transformation is one to one; that is, y =
bution x2 and x ~ Vj. Therefore, Equation 5-30 indicates that the
fils) = pl = py" distribution ofthe random variable Yis
Find the probability distribution of ¥
FA) = A0V3) = PL = PYF", y= 14,9, 16,
‘We now consider the situation where the random variables are continuous. Let ¥ = i(X),
with X continuous and the transformation one to one.
General
Function of | Suppose that X’is a continuous random variable with probability distribution f(x).
8 Continuous | The funetion ¥ = A(X) is a one-to-one transformation between the values of Yand X
Random Variable | so that the equation y = h(x) can be uniquely solved for x in terms of y. Let this
solution be x = u(y). The probability distribution of Y is
Ful») = Flu] LT (5-31)
where J
of Jis used.
(9) is called the Jacobian of the transformation and the absolute value
Equation 5-31 is shown as follows. Let the function y = h(x) be an increasing fimetion of x. Now,
PU Sa) = PIX =< ula)
O
= | pie) de
If we change the variable of integration from x to y by using x = u(y), we obtain dr = wy) dy
and then
rorsa)~ | fluoywone
Since the itera gives the probability hat Y= a forall values ofa contsined inthe Feasible
set of values for y, f[u(y)]u'(y) must be the probability density of Y. Therefore, the probax
Say dneiton of Ps
BO) = lub) = Hl
If the function y = h(x) is a decreasing function of x, a similar argument holds, 55
EXAMPLE 5-35
Let X’be a continuous random variable with probability disti-
bution
fis) =2, Osx<4
x.
Find the probability distribution of Y= A(X) = 24+ 4.
Note that y = h(x) = 2x + 4 isan increasing funtion of
The inverse solution sx = u(y) = (y = 4)/2, and from this
EXERCISES FOR SECTION 5-5
GENERAL FUNCTIONS OF RANDOM VARIABLES 187
Function of a Continuous Random Variable
‘we find the Jacobian to be J = w’(y) = duldy = 1/2. Therefore,
rom Equation 5-51, the probability distribution of Ys
waar)
Say)
5-67. Suppose that isa random vatable with probability
Aistibution
S)= VA x= 123,8
Find the probability distribution of Y= 24+ 1
5-68, Let X’be a binomial random variable with p = 0.25
and n = 3. Find the probability distribution of the random
variable Y= 2
5-69. Suppose that X is a continuous random variable with
probability distribution
Ses)
(a) Find the probability distribution of the random variable
Y=2K+ 10,
() Find the expected value of ¥.
5-70. Suppose that X_has @ uniform probability distribution
fe)
xsl
‘Show thatthe probability distribution of the random variable
Y= ~2 In Xis chi-squared with two degrees of freedom,
5-71, A random variable X has the following probability
distribution:
(@) Find the probability distribution for ¥ = 1%
(b) Find the probability distribution for ¥ = 7.
(6) Find the probability dstibution for ¥ = In.X
5-72. The velocity ofa particle ina gas isa random variable
V with probability distbution
fio) = are v>0
‘where b isa constant that depends on the temperature ofthe
«gs and the mass ofthe parle
{@) Find the value ofthe constant a
(b) The kinetic energy of the particle is W = m¥2/2. Find the
probability distribution of W.
5.73. Suppose that ‘asthe probability distribution
fal, 1sxs2
FFind the probability distribution of the random variable
Y=er
5-74. ‘The random variable X has the probability distri-
bution
fl) = He OK S4
Find the probability distribution of ¥ = (X— 2)
Supplemental Exercises
5-75. Show that the following function satisfies the proper-
ties ofa joint probability mass function:
= > f69)
0 0 1/4
0 ' ys
1 0 18
1 1 1s
2 2 14
Determine the following:
(@) POr< 05, < 15) () PEs 1)
(©) P< 15) (@ PUW> 05, < 15)
(6) Determine £00), EO"), YX), and 0).
(8) Marginal probability distribution of the random vari-
able x
(2) Conditional probability distribution of ¥ given that X = 1
hy EX = 1)
(i) Are Vand Y independent? Why ot why not?
) Calculate the correlation between X and ¥.
5-76. The percentage of people given an anticheumatoid
‘medication who suffer severe, moderate, or minor side effects
are 10, 20, and 70% respectively. Assume that people react 188 CHAPTERS JOINT PROBABILITY DISTRIBUTIONS
independently and that 20 people are given the medication
Determine he following:
(@) The probability that 2,4, and 14 people will suffer severe,
moderate or minor side effects, respectively
(©) The probability that no one will suffer severe side effets
(6) The mean and variance ofthe numberof people who will
sale severe side effets
(@ What is the conditional probability distribution of the
‘numberof people who suffer severe side elects given that
19 suffer minor side effets?
(©) Whats the conditional mean of the number of people who
suler severe side effects given that 19 suffer minor side
effects?
‘The backoff torque required to remove bolts in a ste]
plate is rated as high, moderate, or low. Historically, the prob-
ability ofa igh, moderate, or low rating is 0.6, 0.3, or 0.1,
respectively, Suppose that 20 bolts are evaluated and that the
torque ratings ae independent.
(@) What isthe probability that 12,6, and 2 bolts are rated as
high, moderate, and low, respectively?
(©) What is the marginal distribution ofthe number of bolts
rated low?
(©) What isthe expected number of bolts rated low?
(@) What isthe probebilty thatthe number of bolts rated low
js greater than two?
(©) Whats the conditional distribution ofthe number of bolts
rated low given that 16 bolts are rated high?
(6) What is the conditional expected number of bots rated
low given that 16 bolts ae rated bigh’
(@) Arc the numbers of bolts rated high and low independent
random vatiables?
5-78. To evaluate the technical support from @ computer
‘manufactures, the numberof rings before aeall is answered by
4 service representative is tracked. Historically, 70% of the
calls are answered in two rings or less, 25% are answered in
three or four rings, and the remaining calls require five rings
or more. Suppose you call this manufacturer 10 simes and
‘assume that the ealls are independent.
{(@) What isthe probability that eight calls are answered in two
rings or less, one cal is answered in taree or four rings,
and one call requires five rings or more?
(b) What is the probability that all 10 calls are answered in
four rings or less?
(©) What is the expected number of calls answered in four
rings or less?
(@) What is the conditional distribution ofthe number of calls
requiring five rings or more given that eight calls are
answered in two rings or less?
(©) What isthe conditional expected number of calls requir
ing five rings or more given that eight calls are answered in
two rings or less?
(0) Ave the number of calls answered in two rings of Tess and
the numberof calls requiring five rings or more independ
cent random variables?
5-79. Determine the value of ¢ such that the fanetion
fs»)
‘properties ofa joint probability density funtion,
Determine the following
@Pe<1y<1) &) Pe<25)
(Rls ¥<28) (@) PD 21 <¥<15)
©) EX) ® EY)
(@) Marginal probability distribution of the random variable X
(0) Conditional probability distribution of ¥ given that X = 1
{) Conditional probability distribution of Xgiven that Y= 1
5.80. The joint distribution ofthe continuous random vari-
ables X, Y, and Z is constant over the region 7 + y? = 1.
(0-< 2 < 4, Determine the following:
sey for 0
