=
=
( ) 1 f x dx
=
}
2
  ( ) F x P X x = s
  ( ) ( )
x
F x P X x f x dx
= s =
}
3
  Mean ( )
i i
i
E X x p x = =
  Mean ( ) E X xf x dx
= =
}
4
2 2
( )
i i
i
E X x p x ( =
2 2
( ) E X x f x dx
( =
}
5
( ) ( ) ( )
2
2
Var X E X E X ( =
( ) ( ) ( )
2
2
Var X E X E X ( =
6
Moment =
r r
i i
i
E X x p ( =
Moment = ( )
r r
E X x f x dx
( =
}
7 M.G.F M.G.F
( ) ( )
tX tx
X
x
M t E e e p x ( = =
( ) ( )
tX tx
X
M t E e e f x dx
( = =
}
4) ( ) ( ) E aX b aE X b + = +
5) ( ) ( )
2
Var Var aX b a X + =
6) ( ) ( ) ( )
2 2
Var Var aX bY a X bVar Y = +
7) ( ) Standard Deviation Var X =
8) ( ) ( ) f x F x ' =
9) ( ) 1 ( ) p X a p X a > = s
10) ( )
( )
( )
/
p A B
p A B
p B
= , ( ) 0 p B =
11) If A and B are independent, then ( ) ( ) ( ) p A B p A p B = .
12) 1
st
Moment about origin =   E X =
( )
0
X
t
M t
=
( '
(Mean)
2
nd
Moment about origin =
2
E X (
=
( )
0
X
t
M t
=
( ''
The coefficient of
!
r
t
r
=
r
E X (
(r
th
Moment about the origin)
13) Limitation of M.G.F:
i) A random variable X may have no moments although its m.g.f exists.
ii) A random variable X can have its m.g.f and some or all moments, yet the
m.g.f does not generate the moments.
iii) A random variable X can have all or some moments, but m.g.f does not
exist except perhaps at one point.
14) Properties of M.G.F:
i) If Y = aX + b, then ( ) ( )
bt
Y X
M t e M at = .
ii) ( ) ( )
cX X
M t M ct = , where c is constant.
iii) If X and Y are two independent random variables then
( ) ( ) ( )
X Y X Y
M t M t M t
+
= .
15) P.D.F, M.G.F, Mean and Variance of all the distributions:
Sl.
No.
Distributio
n
P.D.F ( ( ) P X x = )
M.G.F Mean Variance
1 Binomial x n x
x
nc p q
( )
n
t
q pe +
np npq
2 Poisson
!
x
e
x
( )
1
t
e
e
3 Geometric 1 x
q p
(or)
x
q p
1
t
t
pe
qe
1
p
2
q
p
4 Negative
Binomial
1
( 1)
k x
k
x k C p p
+
1
k
t
p
qe
 

\ .
kq
p
2
kq
p
5 Uniform
1
,
( )
0, otherwise
a x b
f x b a
< <
( )
bt at
e e
b a t
2
a b +
2
( )
12
b a
6 Exponential
, 0, 0
( )
0, otherwise
x
e x
f x
> >
=
2
1
7 Gamma 1
( ) , 0 , 0
( )
x
e x
f x x
= < < >
I
1
(1 ) t
8 Weibull
1
( ) , 0, , 0
x
f x x e x

 o
o o 
= > >
16) Memoryless property of exponential distribution
( ) ( ) / P X S t X S P X t > + > = > .
UNITII (RANDOM VARIABLES)
1) 1
ij
i j
p =
= =
}
Marginal density function of Y, ( ) ( ) ( , )
Y
f y f y f x y dx
= =
}
7) ( 1) 1 ( 1) P X Y P X Y + > = + <
8) Correlation co efficient (Discrete):
( , )
( , )
X Y
Cov X Y
x y
o o
=
1
( , ) Cov X Y XY XY
n
=
,
2 2
1
X
X X
n
o =
,
2 2
1
Y
Y Y
n
o =
9) Correlation co efficient (Continuous):
( , )
( , )
X Y
Cov X Y
x y
o o
=
( ) ( ) ( ) ( , ) , Cov X Y E X Y E X E Y = , ( )
X
Var X o = , ( )
Y
Var Y o =
10) If X and Y are uncorrelated random variables, then ( , ) 0 Cov X Y = .
11) ( ) ( ) E X xf x dx
=
}
, ( ) ( ) E Y yf y dy
=
}
, ( ) , ( , ) E X Y xyf x y dxdy
=
} }
.
12) Regression for Discrete random variable:
Regression line X on Y is ( )
xy
x x b y y = ,
( ) ( )
( )
2 xy
x x y y
b
y y
=
Regression line Y on X is ( )
yx
y y b x x = ,
( ) ( )
( )
2 yx
x x y y
b
x x
=
Correlation through the regression, .
XY YX
b b = Note: ( , ) ( , ) x y r x y =
13) Regression for Continuous random variable:
Regression line X on Y is ( ) ( ) ( )
xy
x E x b y E y = ,
x
xy
y
b r
o
o
=
Regression line Y on X is ( ) ( ) ( )
yx
y E y b x E x = ,
y
yx
x
b r
o
o
=
Regression curve X on Y is ( ) ( ) / / x E x y x f x y dx
= =
}
Regression curve Y on X is ( ) ( ) / / y E y x yf y x dy
= =
}
14) Transformation Random Variables:
( ) ( )
Y X
dx
f y f x
dy
= (One dimensional random variable)
( , ) ( , )
UV XY
x x
u v
f u v f x y
y y
u v
c c
c c
=
c c
c c
(Two dimensional random variable)
15) Central limit theorem (Liapounoffs form)
If X1, X2, Xn be a sequence of independent R.Vs with E[Xi] = i and Var(Xi) = i
2
, i
= 1,2,n and if Sn = X1 + X2 + + Xn then under certain general conditions, Sn
follows a normal distribution with mean
1
n
i
i
=
=
and variance
2 2
1
n
i
i
o o
=
=
as
n .
16) Central limit theorem (Lindberg Levys form)
If X1, X2, Xn be a sequence of independent identically distributed R.Vs with E[Xi]
= i and Var(Xi) = i
2
, i = 1,2,n and if Sn = X1 + X2 + + Xn then under certain
general conditions, Sn follows a normal distribution with mean n and variance
2
no as n .
Note:
n
S n
z
n
= ( for n variables),
X
z
n
= = =
Mean   ( ) E X t t = ,
2 2 2
( ) E X t t t ( = +
,   ( ) Var X t t = .
UNITIV (QUEUEING THEORY)
n Number of customers in the system.
Mean arrival rate.
Mean service rate.
n
P Steady State probability of exactly n customers in the system.
q
L Average number of customers in the queue.
s
L Average number of customers in the system.
q
W Average waiting time per customer in the queue.
s
W Average waiting time per customer in the system.
Model I (M / M / 1): ( / FIFO)
1) Server Utilization
=
2) ( ) 1
n
n
P = (P0 no customers in the system)
3)
1
s
L
4)
2
1
q
L
5)
( )
1
1
s
W
=
6)
( ) 1
q
W
=
7) Probability that the waiting time of a customer in the system exceeds t is
( )
( )
t
s
P w t e
> = .
8) Probability that the quue size exceeds t is ( )
1 n
P N n
+
> = where 1 n t = +
.
Model II (M / M / C): ( / FIFO)
1)
s
=
2)
( ) ( )
( )
1
1
0
0
! ! 1
n s
s
n
s s
P
n s
=
(
( = +
3)
( )
( )
1
0 2
1
. !
1
s
q
s
L P
s s
+
=
4)
s q
L L s = +
5)
q
q
L
W
=
6)
s
s
L
W
=
7) The probability that an arrival has to wait: ( )
( )
( )
0
! 1
s
s
P N s P
s
> =
8) The probability that an arrival enters the service without waiting = 1 P(an
arrival hat to wait) = ( ) 1 P N s >
9) ( )
( 1 )
0
( ) 1
1
!(1 )( 1 )
s t s s
t
s e
P w t e P
s s s
(
> = +
`
)
Model III (M / M / 1): (K / FIFO)
1)
=
2)
0 1
1
1
k
P
(No customer)
3) ( )
0
1 P ' = (effective arrival rate)
4)
( )
1
1
1
1 1
k
s k
k
L
+
+
+
=
5)
q s
L L
'
=
6)
s
s
L
W
=
'
7)
q
q
L
W
=
'
8)  
0
a customer turned away
k
k
P P P = =
Model IV (M / M / C): (K / FIFO)
1)
s
=
2)
( ) ( )
1
1
0
0
! !
n s
s k
n s
n n s
s s
P
n s
= =
(
( = +
(
3)
( )
( )
0
0
,
!
,
!
n
n
n
n s
s
P n s
n
P
s
P s n k
s s
s s
4) Effective arrival rate: ( )
1
0
s
n
n
s s n P
=
(
' =
(
5)
( ) ( )
( )
( )
1
0 2
1
! 1
1
s k s
k s
q
s k s
L P
s
+
(
( =
(
6)
s q
L L
'
= +
7)
q
q
L
W
=
'
8)
s
s
L
W
=
'
UNITV (NON MARKOVIAN & QUEUEING NETWORK)
1) Pollaczek Khintchine formula:
( )
 
2
2
( ) ( )
( )
2 1 ( )
S
Var t E t
L E t
E t
(
+
= +
(or)
( )
2 2 2
2 1
S
L
o
+
= +
2) Littles formulas:
( )
2 2 2
2 1
S
L
o
+
= +
q S
L L =
S
S
L
W
=
q
q
L
W
=
3) Series queue (or) Tandem queue:
The balance equation
00 2 01
P P =
1 10 00 2 11
P P P = +
01 2 01 1 10 2 1 b
P P P P + = +
1 11 2 11 01
P P P + =
2 1 1 11 b
P P =
Condition
00 10 01 11 1
1
b
P P P P P + + + + =
4) Open Jackson networks:
i) Jacksons flow balance equation
1
k
j j i ij
i
r P
=
= +
Where k number of nodes, rj customers from outside
ii) Joint steady state probabilities
( ) ( ) ( ) ( )
1 2
1 2 1 1 2 2
, , ... 1 1 ... 1
k
n n n
k k k
P n n n =
iii) Average number of customers in the system
1 2
1 2
...
1 1 1
k
S
k
L
= + + +
iv) Average waiting time of a customers in the system
S
S
L
W
= where
1 2
...
k
r r r = + + +
5) Closed Jackson networks:
In the closed network, there are no customers from outside, therefore 0
j
r =
then
i) The Jacksons flow balance equation
1
k
j i ij
i
P
=
=
0
j
r =
(or)
( ) ( )
11 12 1
2 21 22
1 2 1 2
1 2
...
...
... ...
...
k
k
k k
k k kk
P P P
P P P
P P P
 


=



\ .
ii) If each nodes single server
( )
1 2
1 2 1 2
, , ... ...
k
n n n
k N k
P n n n C =
Where
1 2
1 2
1
1 2
...
...
k
k
n n n
N k
n n n N
C
+ + + =
=
iii) If each nodes has multiple servers
( )
1 2
1 2
1 2
1 2
, , ... ...
k
n n n
k
k N
k
P n n n C
a a a
=
Where
1 2
1 2
1 1 2
...
1 2
...
k
k
n n n
k
N
n n n N
k
C
a a a
+ + + =
=
Prepared by C.Ganesan, M.Sc., M.Phil., (Ph: 9841168917) Page 13
! ,
! ,
i i
i i i
i
n s
i i i i
n n s
a
s s n s
<
=
>
 All the Best 