10/5/2020

LECTURE 6
CIVE 683 – Reliability-Based Design of Civil
Systems

Multiple Random Variables

Fall 2020 Dr. Shadi Najjar AUB

Multiple Random Variable

RV1 RV2

Duration, x Productivity, y Number of Joint PMF

(hr) (%) Observations pX,Y(x,y)
6 50 2 0.014
6 70 5 0.036
6 90 10 0.072
8 50 5 0.036
8 70 30 0.216
8 90 25 0.180
10 50 8 0.058
10 70 25 0.180
10 90 11 0.079
12 50 10 0.072
12 70 6 0.043
12 90 2 0.014
139 1.0

1
 10/5/2020

Multiple Random Variables

Joint PMF

0.25
0.20
0.15
pX,Y(x,y) 0.10
0.05
0.00
90
6 70
8 50
10 Y (%)
X (hr) 12

Multiple Random Variables

Joint Probability Density Function

fX ,Y (x , y)

Joint Cumulative Distribution Function

a b
F X,Y (x , y ) P(X d a ˆ Y d b) ³ ³ f X,Y (x, y )dydx
f f

2
 10/5/2020

Marginal Probability Distributions

Marginal Probability Density Functions
f
f X (x) ³ f X,Y (x, y)dy
f
Marginal PMF

Marginal PMF 0.5

x pX(x)
0.4
6 0.122
8 0.432 0.3
10 0.317 pX(x)
0.2
12 0.129
1.0 0.1

0.0
6
8
10
12
X (hr)

Marginal Probability Distributions

Marginal Probability Density Functions
f
f Y (y ) ³ f X,Y (x, y)dx Marginal PMF
f
0.5

0.4

Marginal PMF 0.3

pY(y)
y p(y) 0.2
50 0.180
70 0.475 0.1

90 0.345 0.0
1.0 50
70
90
Y (%)

3
 10/5/2020

Conditional Probability Distributions

fX ,Y ( x, y)
fX Y (x y) or fX ,Y ( x, y) fX Y ( x y )fY (y )
fY (y )
fX ,Y ( x, y)
fY X (y x) or fX ,Y ( x, y) fY X ( y x )fX ( x)
fX (x )
Conditional PMF

0.5
Conditional PMF

pY|x(y|x = 8 hr)
0.4
y pY|x(y|x=8)
50 0.0833 0.3
70 0.500
0.2
90 0.417
1.0 0.1

0.0
50
70
90
Y (%)

Multiple Random Variables

Covariance and Correlation

COV( X, Y) E[( X  P X )(Y  PY )] E(XY )  E( X)E (Y )

f f
E(XY) ³ ³ xyf X,Y (x , y )dydx
f f

COV( X, Y)
U 1. 0 d U d 1. 0
VXVY

Statistical Independence

fX ,Y (x , y) fX ( x)fY ( y)
COV( X, Y) 0. 0 and U 0. 0

4
 10/5/2020

Multiple Random Variables

Bivariate Normal Distribution

ª 1 §y  P Y  U V Y V X ( x P X ) · ºª
2

¸ º
2
 ¨ 1 §x P X ·
« 1 2 © V Y 1U 2
¸ »
¹ « 1  ¨
2 © V X ¹ »
f X,Y (x , y ) « e »« e
»
« 2SV Y 1  U »¬ 2SV X
2

¬ ¼ ¼

fX ,Y (x , y) fY X (y x) fX (x)

Note: fY X (y x) is normal with

E(Y X x) P Y  U V Y V X (x  P X )

Var (Y X x) V 2Y (1  U2 )

5
 10/5/2020

Bivariate Normal Distribution

PX 0
VX 1
PY 0
VY 1
UX,Y -0.8 0.30
0.25
0.20
fX,Y(x,y) 0.15
0.10 3.
0.05 00
1.
0.00 50
-3.00 0.
00
-1.50 -1

0.00
. 50 y

1.50
x -3
. 00

3.00

Example
Wave height (H) and period (Tp) are modeled using a
bivariate normal distribution:
PH 10 m , V H 2 m , P Tp 15 s , V T p 1. 5 s , U 0. 4

With no information concerning the wave period,

§14  10 ¸
·
P(H ! 14 m ) 1  )¨ 1  )(2.00) 1  0.977 0. 023
© 2 ¹
What if the period is observed: Tp 18 s
E(H T p 18 s) 10  0. 4 2 1.5 (18  15) 11. 6m
Var (H T p 18 s ) 2 2 (1  0. 4 2 ) 3. 36 m 2
§14  11. 6 ·
P(H ! 14 m ) 1  )¨ ¸ 1  )(1.31) 1  0. 905 0. 095
© 3. 36 ¹