Professional Documents
Culture Documents
Event
A set of _________ which is a subset of
(ex) The faces are not same in tossing two coins: { _________ }
P[A B]
P[A|B] =
P[B]
(Em 2.21) Two coins are flipped. What is the prob. of having 2 heads if
at least one is head?
P[D] D
P[B|D]
P[BD]
(Proof)
If A and B are independent, P[A|B] = P[A] -----(a)
P(A B)
By definition, P[A|B] = P(B)
-----------------(b)
(a) = (b) results in
P(A B)
P[A] = P(B)
A1 A3
A5
B
A2
A4
(ex)
12 13 14
4!
21 23 24 43 =
(4 - 2)!
31 32 34
41 42 43
NPR = N(N1)(N(R1))
(Ex 2.13) How many different combinations of cards do we have when we
draw five cards from a deck of cards?
N N!
Combinations: =
R R! (N - R)!
N N
Size of power set: = 2
N
R =0 R
N N N N
(x+y) N = 0 xNy0 + xN-1y1 + 2 xN-2y2 + + N x0yN
1
N N N N N
= + + ... + = ( x + y ) = 2N
N
x =1 and y =1
R=0
R 0 1 N
College of Software # 23 Hee Yong Youn
SungKyunKwan University
Combinations
(Ex 2.14) How many different poker hands do we have if we draw
five cards from a deck of cards?
(Ex 2.15) What is the value of random variable, X, which is the sum
of the dots of two dices rolled?
f(x) = __, x = 0, 2
,x=1
1 2 3 x 0, elsewhere
Since F() = 1, f(x)dx =1
-
(1
i =1
p ) p =
i 1
q (1 q )
i =1
= i 1
(q i 1
q i
)=
i =1
( q 0 q 1 ) + ( q 1 q 2 ) + + ( q k 1 q k ) = 1 q k
F(k) f(k)
1
1 2 3 4 5 k 1 2 3 4 5 k
P[K = k] = (1 p ) p
k
for k = ____, 2,3,
k
F(k) = P[K k] = (1 p ) i p = 1 (1 p ) k +1 for k 0
i=0
(Proof)
Let q = 1 p
k k k
(1
i =0
p ) p = i
q (1
i =0
q )i
= (q i
q
i =0
i +1
)=
(q 0 q1 ) + (q1 q 2 ) + + (q k q k +1 ) = 1 q k +1
: {0i1N-i | i = 0,1,,N}
i =0
i
F(k) f(k)
1
1 2 3 4 5 k 1 2 3 4 5 k
(T) T
k
P[K = k in T] = e
k!
(T) -T
k
k
F(k) = e
i =0 i!
1 2 3 4 5 k 1 2 3 4 5 k
T
P = t = : prob. of a success in t (n >> T)
n
n T T
k n-k
P[K in n] = 1-
k n n
(T) 1- T k n-k
n!
=
k!(n - k)! nk n
1- 1-
n n
k
k!(n - k)! n
1 2 k 1
n k 1(1 )(1 ) (1 )
n (T ) k (1 T ) k ((1 T ) T ) T
n
n n
=
k! n k n n
a na
* lim
n
(1 ) = e
n
1
P[k in T] = lim P[k in n] = (T)k 1. e-T
n k!
(ex)
k b(k; 5,0.2) b(k; 20,0.05) Poisson(k; T=1)
0 0.328 0.359 0.368
1 0.410 0.377 0.368
2 0.205 0.189 0.184
3 0.051 0.060 0.061
t t
(Ex 2.18) What is the prob that a dart lands precisely in the middle
of the dart board?
E[K]
Expected value (average) of r.v. K
Center of mass of the PDF
First moment of r.v. K
3 6 x =
1
(2 6 6) = 4
18
k =1 k =1 dp
d
1
= -p
dp k =1
(1 - p) k
(*
i=0
x i
=
x
, x < 1*)
d 1 d f ( x ) f ( x ) g ( x ) f ( x ) g ( x )
= -p [ - 1] (* = 2
*)
dp p dx g ( x ) ( g ( x ))
-1 1
= -p 2 =
p p
-
x f(x) dx - 2E[X] x f(x) dx + (E[X]) f(x) dx
2 2
- -
E[X 2 ] - 2E[X]E[X] + (E[X]) 2
= E[X 2 ] - (E[X]) 2
= 2 (measure of the spread of the distribution)
k =2
1
= p (1 - p) (k - k) (1 - p) + p + p( 2 - 1)
2 k -2
k =2 p
d2 1
= p (1 - p) (1 - p)k +
2 p
k = 2d p
d2 1 1
= p(1 - p) 2 - (1 - p) - 1 +
d p p p
d 1 1 - 2p 1 2(1 - p) 1 2-p
= p(1 - p) - + 1 + = p(1 - p) - 4 + = + =
p2
p p
2
dp p 2
p p p
2(1 - p) 1 1 2 - 2p + p - 1 1 - p
2 = 2
+ - 2= 2
= 2
p p p p p
6 x
2 = 18 16 = 2
k =0 k!
a k 1 1 a k
ak
ea = k = k k = ae a
k =0 k! a k =0 k! k =1 k!
k
a
E[ K ] = kf ( k ) = k e a = ae a e a = a
k =1 k =1 k!
a k 1
Differentiate both sides on a :e = k a
k =0 k!
a k 2 1 2 a k 1 a k 1 2 a k 1 a 1 2 a k
e a = k ( k 1) = 2 k 2 k = 2 k 2 ae = 2 k a 1e a
k =1 k ! a k =1 k ! a k =1 k ! a k =1 k ! a a k =1 k !
k
a
E[ K 2 ] = k 2 e a = e a (e a + a 1e a )a 2 = a 2 + a
k =1 k!
xy
(ex) R: rainfall
independent
T: temperature
r 40 160 t
200
10 3-6 (t - 40)2
F(r)= r F(t)= 14,400
8
3 2(t - 40)
f(r)= 10-6 r 2 f(t)=
8 14,400
2 (sr 3 (t - 40)2 )
f(r,t) = = s ( 3r 2 (t - 40)2 ) = 6sr 2 (t - 40)
rt t
P[100 r 105 & 50 t 55] = F(105,55) F(100,50)
= 0.00226
College of Software # 56 Hee Yong Youn
SungKyunKwan University
Joint CDF and PDF (example)
(Em 2.38) The first coin determines the number of subsequent flips
such that Head two more flips and Tail one more flip.
T: number of Tails; C: number of flips
Obtain F(c,t).
H (case-1)
H
T (case-2)
H H (case-3)
T
H
T (case-4)
T (case-5)
T (case-6)
F(c,t) t=0 1 2
c =2
3
F(c,t) t=0 1 2
c =2
3
f(c,t) t=0 1 2
c =2
3
(Em 2.42) Find f(2) and f(3) of the coin flipping problem.
2
f(2)= f ( 2, i ) = f(2,0) + f(2,1) + f(2,2) = 0 +1/4 + = 1/2
i =0
f(3)=
f(c,t) t=0 1 2
c =2 0 1/4 1/4
3 1/8 1/4 1/8
f(c=2) = 1/2, f(c=3) = 1/2
f(t|c) t=0 1 2
c =2
3
Unconditioning
f(x,y) is usually _________ to get, while f(xy) and f(y) are not.
(Em 2.45) For the coin flipping problem, obtain f(t = i), (i = 0,1,2).
f(t=0)=ft=0c=2f(2) + ft=0c=3f(3) = 0 x 1/2 + 1/4 x 1/2 = 1/8
f(t=1)=ft=1c=2f(2) + ft=1c=3f(3) = ___________________
f(t=2)=ft=2c=2f(2) + ft=2c=3f(3) = ___________________
t t
Select a register s from a set S, and measure the noise X(t,s) at t.
At t = t1, count the resistor whose noise is smaller than x1, divide it
by |S|. Fx(t1)(x1) = P[X(t1) x1]
Repeat for t2, t3, ...., and get CDF of X(t2), X(t3),...
F(x1,x2) = P[X(t1) x1 and X(t2) x2]
When X(t1,s) = X(t2,s), then the r.v.s are _________ distributed
A process is nth order renewal process if the _____ is the same after
every n events
Residual lifetime (R) (R is a r.v. representing residual lifetime of a
r.v. A)
a( ) a( )
r( - te | te) a( | > te) = =
P[A > te]
te
1 - a(s) ds
0
t = te
a(t + te)
r(t|te) = t e
1 - a(s) ds
0
2
f
r = (f is original density)
2f
f ( t ) = e -t
e ( t + t )
e
e t e t e
e t e t e
r (t | te ) = = = = e t
1 (e ) t t e
te te
e
1 e t dt 0
(b) Provided that no job arrives in t =10, what is the prob. of a job
arrival in t =11?
P(T 2| 1 < T) = P(1 < T) = 1 - (1 - e )
2
P(1 < T 2)
- 2t
2e dt 1
-2
2
- e -2t 1 e - 2 - e -4
= = = 1 - e -2
e -2 e -2
e- t (t)0
(Ans) P[X > t] = P[N = 0] = = e- t
0!
k!
1! 1! 1! = h1h2 hk
prob = ( t )k e t tk
k!
2 - 1
2 - t 1 - t
F(t) = 1 e +
1 2
e , t0
2 - 1 2 - 1
k
F(t) = i (1 - e - t
) i
i =1
f(x) = 1 -
e 2 , - < x < , : mean
2
X~N(, 2)
(ex) errors of measurement
No closed form F(x); use table for Z~N(0,1) (standard normal dist)
t2
FZ(z) = 1 e dt
z
2
x
FZ ( z) = 1 FZ (z); FX(x) = FZ ( )
(ex) N(200,256) signal is received. What is the prob that the signal is
greater than 240mV?
240 200
(Ans) P[X> 240] = 1 P[X 240] = 1 FZ ( 16 ) = 1 FZ (2.5)
= 0.0062
f(t) = t e ( 1 ) t
F(t) = 1 e t
Fault modeling
Exponential dist is a special case of Weibull dist with = 1