Добавил:
Upload Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Lektsii (1) / Lecture 10

.pdf
Скачиваний:
13
Добавлен:
02.06.2015
Размер:
90.2 Кб
Скачать

ICEF, 2012/2013 STATISTICS 1 year LECTURES

Lecture 10

13.11.2012

BERNOULLI SCHEME. BINOMIAL RANDOM VARIABLES

Suppose there is a trial with two random outcomes, that we will call success or a failure. Examples: 1) a coin is tossed one time, success = head, failure = tail; 2) a detail is selected from a big lot, success = selected detail is not defective, failure = selected detail is defective; 3) a chip is tested, a chip works without breakdowns at least one month, failure = less than one month, etc. Now let’s assume that we independently make n such trials and the probability of success in each trial is π, 0 π 1 independently on the number of a trial. This series of trials is called

Bernoulli scheme.

Definition. The total number of successes in Bernoulli scheme with n trials and probability of success π is called the binomial random variable with parameters (n,π) .

Let’s denote by 1 the appearance of success and by 0 the appearance of failure. Then the result of any Bernoulli scheme with parameters (n,π) can be represented as a sequence

s = (ε1,ε2 ,...,εn ) ,

 

there is success in i

th

trial,

i =1,..., n.

where εi = 1, if

 

0, if

there is failure in ithtrial,

 

It is clear that the total number of successes X is equal to n εi .

i=1

Let’s assign to each s = (ε1,ε2 ,...,εn ) the probability Pr(s) = pk qnk

where k = n

εi

is the total number of successes. For the fixed value of k there are exactly

 

 

 

i=1

 

 

 

n

 

n!

 

 

 

 

 

=

 

 

outcomes with the total number of successes equal to k. Finally we get:

 

 

 

 

 

 

k!(n k)!

 

k

 

 

n

 

n!

 

Distribution of binomial random variable: Pr(X = k) =

 

=

, k =0,1,...,n .

 

 

 

 

 

k!(n k)!

 

k

 

 

EXPECTATION AND VARIANCE OF A DISCRETE RANDOM VARIABLE Let X be some discrete random variable,

X

x1

x2

xn

P(X)

p1

p2

pn

Definition. Number E(X ) = µX = xi pi is called the expectation or mean value

i

(математическое ожидание) of a random variable X.

Properties of expectation

1. If X 0 then E(X ) 0 (monotonicity).

2.If X, Y are two random variables and a, b are constants then E(aX +bY ) = aE(X ) +bE(Y ) (linearity of expectation)

Definition. Number V (X ) =σX2 = (xi µX )2 pi the variance (дисперсия) of a random

i

variable X. Number V (X ) =σX is called the standard deviation (стандартное отклонение) of a random variable X.

Properties of variance

1.V (X ) 0 for any random variable X, and V (X ) = 0 if and only if random variable is constant (i.e. is not random).

2.If a and b are two constants then V (aX +b) = a2V (X )

3.V (X ) = E(X 2 ) [E(X )]2 = E(X 2 ) µX2 .

Let X be a binomial random variable with parameters (n,π) (recall that n is the number of trials and π is a probability of success in each trial. According to the definition

n

n

πk (1π)nk

E(X ) = k

k =0

k

 

and variance is

n

n

 

V (X ) = (k np)2 πk (1π)nk .

k =0

 

k

Using a routine arithmetic one can prove that

E(X ) = nπ, V (X ) = nπ(1π)

Example. The proportion of defective chips is 60%. Ten chips are selected at random from the large lot.

10

1) Probability that exactly half of chips is defective: P(X =5) = 0.650.45 = 0.20 .

5

2)The expectation of defective chips is E(X ) =10*0.6 = 6 .

3)The variance: V (X ) =σ2 =10*0.6*0.4 = 2.4, σ =1.55 .

Соседние файлы в папке Lektsii (1)