Академический Документы
Профессиональный Документы
Культура Документы
Distributions
Part One: Discrete Distributions
Ch. 3-1
So far...
Uncertainty was formalised as Probability
Theory.
Working with Events too Clumsy
Ch. 3-2
Ch. 3-3
Continuous
Probability
Distributions
Binomial
Uniform
Hypergeometric
Normal
Poisson
4
Ch. 3-5
Binomial Distribution
Used extensively.
Applied to generate the probability
distribution of observing X successes from N
trialsl
Will build up in three steps:
1. Introduce the Bernouilli distribution to
describe the probability distribution for a
single event.
Copyright 2010 Pearson
Education, Inc. Publishing as
Prentice Hall
Ch. 3-6
Binomial Distribution II
2. Review expression for Combinations so to
extend probability for single event to the
number of successes from a larger number of
trials.
3. Put the two together to produce the Binomial
Distribution equation.
Ch. 3-7
Bernoulli Distribution
Random experiment with only two possible outcomes:
Consider only two outcomes: success or failure
Let P denote the probability of success
Let 1 P be the probability of failure
Define random variable X:
x = 1 if success, x = 0 if failure
Then the Bernoulli probability function is
Bernoulli Distribution
Mean and Variance
The mean is = P
Proof:
Sequences of x Successes
in n Trials
The number of sequences with x successes in n independent
trials is:
n!
C =
x! (n x)!
n
x
11
Generalising
Examples:
The number of heads thrown when tossing a coin 100 times
The number of students that show up in a tutorial
The number of correct answers in a test
In principle, we could work out the probabilities corresponding
to a binomial distribution in this way
However, this would quickly become impractical
Suppose the student applies to 20 colleges. What is the
probability that exactly 8 colleges admit her?
We can use a formula that generates the probabilities
12
Generalising
If
A random experiment can have only two outcomes
The probability of a successful trial (p) is constant across
experiments
Each trial is independent of the rest
The random variable X defines the number of successes in n
independent trials with a probability p of success in each trial
13
Generalising
Binomial Distribution Formula
n
P ( X = x) = P( x) = Cx p x (1 p) n x
n!
where C x =
x!(n x)!
n
P( X = x) = Cx p x (1 p) n x
15
n!
P(x = 1) =
P X (1 P) n X
x! (n x)!
= ??
16
Binomial Distribution
The shape of the binomial distribution depends on the values
of P and n
n = 5 P = 0.1
Mean
P(x)
.6
.4
.2
0
x
0
P(x) n = 5 P = 0.5
.6
.4
.2
0
x
0
Binomial Distribution
Mean and Variance
Mean
= E(x) = nP
2 = nP(1- P)
= nP(1- P)
Where n = sample size
P = probability of success
(1 P) = probability of failure
18
Binomial Characteristics
Examples
Mean = nP = ??
= nP(1 - P) = ??
n = 5 P = 0.1
P(x)
.6
.4
.2
0
x
0
= nP = ??
= nP(1 - P) = ??
P(x) n = 5 P = 0.5
.6
.4
.2
0
x
0
Example (3)
1) Bits are sent over a communications channel in packets of 12. If
the probability of a bit being corrupted over this channel is 0.1 and
such errors are independent, what is the probability that no more
than 2 bits in a packet are corrupted?
2) If 6 packets are sent over the channel, what is the probability that
at least one packet will contain 3 or more corrupted bits?
3) Let X denote the number of packets containing 3 or more
corrupted bits. What is the probability that X will exceed its mean
by more than 2 standard deviations?
20
Other Distributions:
The Hypergeometric Distribution
n trials in a sample taken from a finite population of size N
Sample taken without replacement
Outcomes of trials are dependent
Concerned with finding the probability of X successes in the
sample where there are S successes in the population
21
Hypergeometric Distribution
Formula
CSxCnNxS
P(x) =
CnN
S!
(N S)!
Using the
Hypergeometric Distribution
Example: 3 different computers are checked from 10 in the
department. 4 of the 10 computers have illegal software
loaded. What is the probability that 2 of the 3 selected
computers have illegal software loaded?
23
Poisson Distribution
In many practical situations we are interested in measuring how
many times a certain event occurs in a specific time interval or in a
specific length or area. For instance:
The number of phone calls received at an exchange or call centre
in an hour;
The number of customers arriving at a toll booth per day;
The number of flaws on a length of cable;
The number of cars passing using a stretch of road during a day.
The Poisson distribution plays a key role in modelling such problems.
24
Poisson Distribution
Suppose we are given an interval (this could be time, length, area or
volume) and we are interested in the number of successes in that
interval.
Assume that the interval can be divided into very small subintervals
such that:
The probability that an event occurs in one subinterval is very
small
The probability of one success in a subinterval is constant for all
subintervals and is proportional to its length;
Subintervals are independent of each other.
25
Poisson Distribution
Formula
We assume the following:
The random variable X denotes the number of successes in
the whole interval.
is the mean number of successes in the interval.
X has a Poisson Distribution with parameter and
e
P(x) =
x!
where:
x = number of successes per unit
= expected number of successes per unit
e = base of the natural logarithm system (e=exp(1)=2.71828...)
26
p (x )
0.000002
0.000027
0.000175
0.000766
0.002510
0.006575
0.014356
0.026866
0.043994
0.064036
x
10
11
12
13
14
15
16
17
18
19
p (x )
0.083887
0.099901
0.109059
0.109898
0.102833
0.089807
0.073530
0.056661
0.041237
0.028432
P [ a t le a st 1 0 ] = P [ X 1 0 ] = 1 P [ X 9 ]
P a t m o s t 1 5 a t le a s t 1 0 = P X 1 5 X 1 0
30
31
32
33
34
35
V (X) = .
36
Poisson Approximation to
Binomial Distribution
The Poisson distribution can be derived as a limiting case to the
binomial distribution as 1) the number of trials goes to infinity 2)
the expected number of successes remains fixed
1) n
2) p 0 with np staying constant
Then, writing := np, it can be shown that the binomial
probabilities bin(k; n, p) tend to the Poisson probability
Poisson probabilities can be used to approximate binomial
probabilities when n is large and p is small.
As a rule of thumb, this approximation is acceptable if n 20 and p
0.05 If n 100 and np 10, it is usually an excellent
37
approximation.
Poisson Approximation to
Binomial Distribution
Example: It is known that 3% of the circuit boards from a
production line are defective. If a random sample of 120 circuit
boards is taken from this production line, use the Poisson
approximation to estimate the probability that the sample contains:
(i) Exactly 2 defective boards.
38
Poisson Approximation to
Binomial Distribution
Example: It is known that 3% of the circuit boards from a
production line are defective. If a random sample of 120 circuit
boards is taken from this production line, use the Poisson
approximation to estimate the probability that the sample contains:
(ii) At least 2 defective boards.
39