Binomial and Poission Probablity distribution
Download as PPTX, PDF49 likes23,600 views
The document discusses binomial and Poisson distributions. Binomial distribution describes random events with two possible outcomes, like success/failure. Poisson distribution models rare, independent events occurring randomly over an interval of time/space. An example calculates the probability of defective thermometers using binomial distribution. It also fits a Poisson distribution to automobile accident data from a 50-day period.
Binomial and Poission Probablity distribution
- 1. Binomial and Poisson Distribution By Prateek Singla & Avtar Singh
- 2. What is a Probability Distribution? Listing of all possible outcomes of an experiment and probability of happening each of those events. Characteristics of a Probability distribution- Probability of a particular outcome can be between 0 and 1, both inclusive. The outcomes are mutually exclusive events. The list is completely exhaustive. So, sum of all probabilities of various events would always be 1.
- 3. Binomial Distribution It is a discrete probability distribution. Binomial Probability is calculated by following general formula- P(X) = n Cx px q(n-x) Where, n = number of trials x = number of success p = Probability of success q = Probability of failure = 1 – p
- 4. Requirements for a Binomial Distribution Random Experiment must involve n identical trials. As the word “Binomial” suggests, each trail should have only 2 possible outcomes, denoted as “Success” or “Failure”. Each trial is independent of the previous trials. The probability of success denoted by p, does not change from trial to trial. The probability of failure is 1-p and it is also fixed from trial to trial.
- 5. A Simple Managerial Problem ABS Laboratories is making a thermometer with 20% defect rate. If we select 5 randomly chosen thermometers at end of manufacturing process, what is the probability of having exactly 1 defective thermometer in our sample? Hint – The answer is NOT 1/5 or 0.2 or 20 %
- 6. A Simple Managerial Problem Trail 1 2 3 4 5 Outcome F F S F F Outcome S F F F F Outcome F F F S F Outcome F S F F F Outcome F F F F S Here is a possible list of outcomes-
- 7. A Simple Managerial Problem Trail 1 2 3 4 5 Probability Outcome 0.8 0.8 0.2 0.8 0.8 0.08192 Outcome 0.2 0.8 0.8 0.8 0.8 0.08192 Outcome 0.8 0.8 0.8 0.2 0.8 0.08192 Outcome 0.8 0.2 0.8 0.8 0.8 0.08192 Outcome 0.8 0.8 0.8 0.8 0.2 0.08192 Probability of success (Thermometer being defective) = 0.2 Probability of failure = 1 – 0.2 = 0.8 0.4096
- 8. A Simple Managerial Problem It is given in problem that number of randomly selected samples is 5, therefore, n = 5 It is also given that probability of a thermometer being defective is 20% or 0.2, therefore- p = 0.2 , q = 1 – p = 1 – 0.2 =0.8 We have to find probability of exactly one being defective, therefore x = 1, Using formula- P(X) = n Cx px q(n-x) » P(1) = 5 C1 (0.2) 1 (0.8) (5-1) = 0.4096 Hence, probability of finding exactly 1 defective thermometer out of 5 is approximately 41%
- 9. Tabulating Probability Distribution n #successes #failures Calculation Probability Defective 5 0 5 5 C0 x (0.2) 0 x (0.8) 5 0.32768 0 5 1 4 5 C1 x (0.2) 1 x (0.8) 4 0.4096 1 5 2 3 5 C2 x (0.2) 2 x (0.8) 3 0.2048 2 5 3 2 5 C3 x (0.2) 3 x (0.8) 2 0.0512 3 5 4 1 5 C4 x (0.2) 4 x (0.8) 1 0.0064 4 5 5 0 5 C5 x (0.2) 5 x (0.8) 0 0.00032 5 Calculation for all possible cases of success/failure. 1
- 10. Using Binomial Table n = 5 x 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.590 0.328 0.168 0.078 0.031 0.010 0.002 0.000 0.000 1 0.328 0.410 0.360 0.259 0.156 0.077 0.028 0.006 0.000 2 0.073 0.205 0.309 0.346 0.313 0.230 0.132 0.051 0.008 3 0.008 0.051 0.132 0.230 0.313 0.346 0.309 0.205 0.073 4 0.000 0.006 0.028 0.077 0.156 0.259 0.360 0.410 0.328 5 0.000 0.000 0.002 0.010 0.031 0.078 0.168 0.328 0.590 Page 781 of Book
- 11. Graphing Probability Distribution Zero One Two Three Four Five Defective Thermometer 0.32768 0.4096 0.2048 0.0512 0.0064 0.00032 0.32768 0.4096 0.2048 0.0512 0.0064 0.00032 Probability of Finding Defective Thermometer Defective Thermometer
- 12. Mean of Binomial Distribution Mean or expected value for Binomial distribution is given as- µ = np Therefore, µ = 5 x 0.2 Mean = 1 n p q 5 0.2 0.8
- 13. Variance of Binomial Distribution Variance of Binomial Distribution is given as- σ2 = npq = 5 x 0.2 x 0.8 = 0.8 n p q 5 0.2 0.8
- 14. Standard Deviation of Binomial Distribution n p q 5 0.2 0.8
- 15. Final Results of Problem n p q 5 0.2 0.8 Probability of finding one defective thermometer 40.96 % Mean Variance Standard Deviation 1 0.8 0.8944
- 16. Problem on Binomial Fitting The following data shows the number of seeds germinating out of 12 on a damp filter for 128 set of seeds. Fit a binomial distribution and find the expected frequencies if the chance of seed germination is 0.5 Also, find mean and standard deviation for the same. No. of germinations 0 1 2 3 4 5 6 7 Total No. of samples 7 6 19 35 30 23 7 1 128
- 17. Binomial Fitting p q N 0.5 0.5 128 Since there are 8 terms, therefore, n = 7 Expected frequencies are nothing but coefficients of binomial expansion of N(p+q) n . Hence, a generalised formula for expected frequency is- fx = n Cx
- 18. Calculating mean, variance and standard deviation
- 20. Discovered by Mathematician Simeon Poisson in France in 1781 The modeling distribution that takes his name was originally derived as an approximation to the binomial distribution. HISTORICAL NOTE
- 21. POISSON EXPERIMENT PROPERTIES • Counting the number of times a success occur in an interval • Probability of success the same for all intervals of equal size • Number of successes in interval independent of number of successes in other intervals • Probability of success is proportional to the size of the interval • Intervals do not overlap
- 22. THE POISSON EXPERIMENT • Used to find the probability of a rare event • Randomly occurring in a specified interval – Time – Distance – Volume • Measure number of rare events occurred in the specified interval
- 23. ROLE OF POISSON DISTRIBUTION • It is used in quality control statistics to count the numbers of defects of any item • In biology to count the number of bacteria • In physics to count the number of particles emitted from a radio active substance • In insurance problems to count the number of casualities • In waiting time problems to count the number of incoming telephone number of incoming telephone calls or incoming customers • The number of typographical errors per page in typed material
- 24. Calculating the Poisson Probability Determining x successes in the interval µ = mean number of successes in interval e = base of natural logarithm = 2.71828
- 25. The Poisson Experiment Example On average the anti-virus program detects 2 viruses per week on a notebook • Time interval of one week • μ = 2 per week • Occurrence of viruses are independent • yes we Can calculate the probabilities of a certain number of viruses in the interval Are the conditions required for the Poisson experiment met?
- 26. Poisson Experiment Example Let X be the Poisson random variable indicating the number of viruses found on a notebook per week P ( X = x ) = P ( x ) = µx e- µ /x! P ( x = 0 ) = P ( 0 ) = 20 e- 2 /0! = 0.1353 P ( x = 1 ) = P ( 1 ) = 21 e- 2 /1! = 0.2707 P ( x = 3 ) = P ( 3 ) = 23 e- 2 /3! = 0.1804 P ( x = 7 ) = P ( 7 ) = 27 e- 2 /7! = 0.0034 X P ( x ) 0 0.1353 1 0.2707 2 0.2707 3 0.1804 7 0.0034 ∑p ( x ) = 1
- 27. Calculate the probability that two or less than two viruses will be found per week P ( x ≤ 2 ) = p ( x = 0 ) + p ( x =1 ) = p ( x = 2 ) = 0.1353 + 0.2707 + 0.2707 = 0.6767 The Poisson Experiment Example
- 28. Calculate the probability that less than three viruses will be found per week P ( x < 3 ) = p ( x = 0 ) + p ( x =1 ) = p ( x = 2 ) = 0.1353 + 0.2707 + 0.2707 = 0.6767 The Poisson Experiment Example
- 29. Calculate the probability that more than three viruses will be found per week P ( x > 3 ) = p ( x=4 ) + p ( x = 5 ) + ……… = 1- p ( x ≤ 3 ) 1 – 0.8571 0.1429 The Poisson Experiment Example
- 30. Calculate the probability that four viruses will be found in four weeks μ = 2 x 4 = 8 in four weeks P ( x = 4 ) = 84 e- 8 /4! = 0.0573 The Poisson Experiment Example
- 31. The Poisson Experiment Example • Mean and standard deviation of Poisson random variable
- 32. Fitting a Poisson distribution The following tables gives a number of days in a 50 – day period during which automobile accidents occurred in a city Fit a Poisson distribution to the data : No of accidents : 0 1 2 3 4 No. of days :21 18 7 3 1 solution X F Fx 0 21 0 1 18 18 2 7 14 3 3 9 4 1 4 N=50 ∑fx=45 Fitting of Poisson distribution M = X = ∑fx/N = 45/50 = 0.9
- 33. Fitting a Poisson distribution f ( 0 ) = e-m = e-0.9 = 0.4066 f ( 1 ) = mf ( 0 ) = ( 0.9 ) ( 0.4066 ) = 0.3659 f ( 2 ) = m/2 f ( 1 ) = 0.9/2 ( 0.3659 ) = 0.1647 f ( 3 ) = m/3 f ( 2 ) = 0.9/3 ( 0.1647 ) = 0.0494 f ( 4 ) = m/4 f ( 3 ) = 0.9/4 ( 0.0494 ) = 0.0111 In order to fit Poisson distribution , we shall multiply each probability by N i.e. 50 Hence the expected frequencies are : X 0 1 2 3 4 F 0.4066 X 50 0.3659 X 50 0.1647 X 50 0.0494 X 50 0.0111 X 50 =20.33 =18.30 =8.24 =2.47 =0.56