Discrete Probability Distributions 1 Probability Basics 2 Probability Basics • Random Experiment • Sample Space, Ω • Event • Intersection, ๐ด ∩ ๐ต • Mutually exclusive, ๐ด ∩ ๐ต = ∅ • Union, ๐ด ∪ ๐ต • Complement, ๐ด๐ถ 3 Probability Basics Probability Postulates 1. If ๐ด is any event in the sample space,Ω, 0≤โ ๐ด ≤1 2. Let ๐ด ⊆ Ω and let ๐๐ denote the basic outcomes. Then โ ๐ด = เท โ ๐๐ ๐ด 3. โ Ω = 1. 4 Probability Basics • Complement Rule: โ ๐ด๐ = 1 − โ ๐ด . • Addition Rule: โ ๐ด ∪ ๐ต = โ ๐ด + โ ๐ต − โ ๐ด ∩ ๐ต • Conditional Probability: โ ๐ด|๐ต โ ๐ด∩๐ต = โ๐ต ; โ ๐ต ≠0 • The Multiplication Rule: โ ๐ด ∩ ๐ต = โ ๐ด|๐ต โ ๐ต 5 Probability Basics Two events, ๐ด be and ๐ต,c are said to be statistically independent if and only if โ ๐ด∩๐ต =โ ๐ด โ ๐ต . 6 Probability Basics Bayes’ Theorem Let ๐ด and ๐ต be two events. Then Bayes’ theorem states that โ ๐ด|๐ต โ ๐ต โ ๐ต|๐ด = . โ ๐ด 7 Discrete Random Variables Probability Distributions 8 Random Variables There are different types of variables we will consider: 9 Random Variables Random Variables • Discrete Random Variable • Continuous Random Variables Probability Distribution Function (pdf) 1. 0 ≤ โ ๐ = ๐ฅ ≤ 1 ∀ x ∈ ๐๐ 2. σ๐ฅ∈๐๐ โ ๐ = ๐ฅ = 1 10 Discrete Random Variables Cumulative Probability Distribution (cdf) ๐น๐ ๐ฅ0 = เท โ ๐ = ๐ฅ ๐ฅ≤๐ฅ0 1. 0 ≤ ๐น๐ ๐ฅ0 ≤ 1 ∀ ๐ฅ0 ∈ โ. 2. If ๐ฅ0 < ๐ฅ1 , then ๐น๐ ๐ฅ0 ≤ ๐น๐ ๐ฅ1 . 11 Discrete Random Variables Expected Value ๐ผ ๐ = ๐ = เท ๐ฅโ ๐ = ๐ฅ ๐ฅ∈๐๐ฅ Variance ๐2 = ๐ผ ๐ − ๐ 2 = ๐ผ ๐ 2 − ๐2 = เท ๐ฅ 2 โ ๐ = ๐ฅ − ๐2 ๐ฅ∈๐๐ The standard deviation, ๐ = ๐ 2 12 Discrete Random Variables 13 Discrete Random Variables Let ๐ be a random variable with mean ๐๐ and variance ๐๐2 , and let ๐ and ๐ be any constant fixed numbers. Define the random variable ๐ = ๐ + ๐๐. Then, the mean and variance of Y are ๐๐ ๐๐2 = ๐ + ๐๐๐ = ๐2 ๐๐2 14 Discrete Random Variables 15 Multivariate Basics Let ๐1 , … , ๐๐ be random variables, • ๐ผ ๐1 + โฏ + ๐๐ = ๐ผ ๐1 + โฏ + ๐ผ ๐๐ • ๐ ๐๐ + ๐๐ = ๐ ๐๐ + ๐ ๐๐ + 2Cov ๐๐ , ๐๐ • ๐ σ๐๐=1 ๐๐ = σ๐๐=1 ๐ ๐๐ + 2 σ1≤๐<๐≤๐ Cov ๐๐ , ๐๐ • Cov ๐๐ , ๐๐ = ๐ ๐๐ 16 Multivariate Basics Double conditional expectation Let ๐ and ๐ be two random variables. Then ๐ผ ๐ผ ๐|๐ =๐ผ ๐ 17 Binomial Distribution 18 Binomial Distribution • Two possible mutually exclusive and collectively exhaustive outcomes. • Let ๐ denote the probability of success, and the probability of failure (1 − ๐). 1 if success • ๐=แ 0 otherwise This distribution is known as the Bernoulli distribution. 19 Binomial Distribution Binomial distribution: • Two possible outcomes for each trial. • Mutually exclusive. • Counting successes. • Independent trials. • Identical probability of success for all trials. 20 Binomial Distribution Let ๐ ∼ Bin ๐, ๐ , then โ ๐=๐ฅ ๐ผ ๐ ๐2 = = = ๐ = ๐ฅ ๐−๐ฅ ๐ถ ๐ 1 − ๐ ๐ ๐ฅ ๐๐ ๐๐ 1 − ๐ ๐2 21 Binomial Distribution 22 Binomial Distribution 23 Binomial Distribution 24 Exercises For each of the following, indicate if a discrete or a continuous random variable provides the best definition: a) The amount of oil exported by Saudi Arabia in January 2019. b) The number of newspapers published by The Copenhagen Post. c) The number of rainy days in July at a beach resort. d) The level of pressure in the tires of an automobile. 25 Exercises • A presidential election poll contacts 2,000 randomly selected people. Should the number of people that support candidate A be analyzed using discrete or continuous probability models? • The Embassy of Vietnam in India receives a number of visa applications. Should the number of visas issued per day be analyzed using discrete or continuous probability models? 26 Exercises Show the probability distribution function of the number of heads when three fair coins are tossed independently. 27 Exercises The number of computers sold per day at Dan’s Computer Works is defined by the following probability distribution: ๐ 0 1 2 3 4 5 6 โ ๐=๐ฅ 0.03 0.11 0.15 0.22 0.19 0.26 0.04 a) โ 3 ≤ ๐ฅ < 6 = ? b) โ(๐ฅ > 3) = ? c) ๐(๐ฅ ≤ 4) = ? d) ๐(2 < ๐ฅ ≤ 5) = ? 28 Exercises Given the probability distribution function: ๐ 0 1 2 โ ๐=๐ฅ 0.25 0.50 0.25 a) Graph the probability distribution function. b) Calculate and graph the cumulative probability distribution. c) Find the mean of the random variable ๐. d) Find the variance of ๐. 29 Exercises An automobile dealer calculates the proportion of new cars sold that have been returned a various numbers of times for the correction of defects during the warranty period. The results are shown in the following table. ๐ 0 1 2 3 4 โ ๐=๐ฅ 0.28 0.36 0.23 0.09 0.04 a) Graph the probability distribution function. b) Calculate and graph the cumulative probability distribution. c) Find the mean number of returns of an automobile for corrections for defects during the warranty period. d) Find the variance of the number of returns of an automobile for corrections for defects during the warranty period. 30 Exercises A production manager knows that 5% of components produced by a particular manufacturing process have some defect. Six of these components, whose characteristics can be assumed to be independent of each other, are examined. a) What is the probability that none of these components has a defect? b) What is the probability that one of these components has a defect? c) What is the probability that at least two of these components have a defect? 31 Exercises A local car dealer in Germany is mounting a new cross-divisional promotional campaign. Purchasers of new cars may, if dissatisfied for any reason, return the vehicle within 2 days of purchase and receive a full refund. The cost to the dealer of such a refund is €300. The dealer estimates that 18% of all purchasers will indeed return cars purchased and obtain refunds. Suppose that 80 cars are purchased during the campaign period. a) Find the mean and standard deviation of the number of these cars that will be returned for refunds. b) Find the mean and standard deviation of the total refund costs that will accrue as a result of these 80 purchases. 32