Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
1 Tree Diagrams Example: A man takes either a bus or the subway to work with probabilities 0.3 and 0.7 , respectively. When he takes the bus, he is late 30% of the days. When he takes the subway, he is late 20% of the days. What is the probability that he is late? Solution: First we can write a tree diagram for the corresponding probabilities as follows: 0.3 −→ The man → 0.7 −→ 0.3 −→ by bus → 0.7 −→ 0.2 −→ by subway → 0.8 −→ late on time late on time From above tree diagram, we can get P({The man is late}) = 0.3(0.3) + 0.7(0.2) = 0.23. Example: As items come to the end of a production line, an inspector chooses items to undergo a complete 2 inspection. Of all items produced 20% are defective. 50% of all defective items go through a complete inspection, and 30% of all good items go through a complete inspection. Given that an item is completely inspected, what is the probability that it is defective? Solution: Define D : An item defective. C : An item is completely inspected. Then, we have following information: P(C|D) = 50% P(D) = 20% P(C|Dc) = 30% P(Dc) = 80% 0.5 −→ inspected 0.2 −→ defective → 0.5 −→ not inspected An item → 0.8 −→ 0.3 −→ good → 0.7 −→ inspected not inspected 3 From above tree diagram, we can get we have P(D ∩ C) P(C) P(D)P(C|D) = P(C) 0.2(0.5) 10 = = 0.2(0.5) + 0.8(0.3) 34 P(D|C) = Chapter Thirteen: Binomial Distributions First, let us call the counting rule for combinations. Counting Rule for Combinations The number of different combinations of n different objects that can be formed, taking them r at a time, is n n! = Crn = r!(n − r)! r Example: We want to choose 5 people from 20 people to organize a traveling group. Are there how many different ways to choose a group? Solution: This is just how many different combinations. So the solution is 20 20! = = 15504. 5!(20 − 5)! 5 An experiment is called a binomial experiment if it satisfies following conditions: 4 (1) There are a fixed number n of trials or subexperiments. (2) The n trials are all independent. (3) Each trial has exactly two possible outcomes: success and failure. (4) The probability of success is p which satisfies 0 < p < 1 and which is the same for all the trials. Binomial distribution: Let X denote the count of successes in the binomial experiment. Then X has following binomial distribution: The distribution table of r.v X x 0 1 2 ··· n P(X = k) C0np0q n C1np1q n−1 C2np2q n−2 · · · Cnnpnq 0 where q = 1−p.pThen, X has mean µ = np and standard deviation σ = np(1 − p). Remark: If we consider that n different positions, each time we choose k different positions to put successes. Then, this is just the total number of different combinations we can get and this is where the coefficient Ckn comes from. Example: A couple plans to have three children. There are 8 possible arrangements of girls and boys. For example, GGB means the first two children are girls and the third child is a boy. All 8 arrangements are equally likely. 5 (a) Write down all 8 arrangements of the sexes of three children. What is the probability of any one these arrangements. (b) Let X be the number of girls the couple has. What is the probability that X = 2? (c) Find the distribution of X. Solution: (a) All different 8 arrangements are as follows: 1 2 3 4 5 6 7 8 BBB BBG BGB BGG GBB GBG GGB GGG They are equally likely. So each has probability 1/8. Since X is the number of girls the couple has, P(X = 2) = P({BGG, GBG, GGB}) = P({BGG}) + P({GBG}) + P({GGB}) = 3/8. (c) Similarly we can find that P(X = 1) = P({BBG, BGB, GBB}) = P({BBG}) + P({BGB}) + P({GBB}) = 3/8, P(X = 0) = P({BBB}) = 1/8, and P(X = 3) = P({GGG}) = 1/8. Thus, we have the distribution table of X. 6 Value of X 0 1 2 3 Probability 1/8 3/8 3/8 1/8 Now if we use binomial distribution, then, p = 1/2 and n = 3, we can get 1 1 P(X = 1) = C13( )1( )2 = 3/8 2 2 and Value of X 0 1 2 3 Probability 1/8 3/8 3/8 1/8 Example: A couple plans to have 5 children. Each new baby has equal probability to be a boy or girl. Let X be the number of girls the couple will have. What is the probability that X = 2? Solution: 1 1 P(X = 2) = C25( )2( )3 = 5/16. 2 2 Example: Consider an experiment of randomly drawing a chip 8 times successively with replacement from a box containing two black chips and three white chips. The chips are identical except their colors. Let X be the number of black chips observed in the 8 drawings. (a) Find the probability P(X = 5). (b) Find its mean and variance. Solution: (a) According to the binomial distribution 7 formula, we have P(X = 5) = 5 3 8−5 8 2 C5 ( ) ( ) 5 5 = 0.124. (b) Its mean is equal to µ = 8(2/5) = 16/5 and its variance is equal to 48 25 Example: Consider a binomial experiment of flipping a biased coin twenty times successively. Let A be the event of observing a head and Ā be the event of observing a tail. Suppose that σ 2 = 8(2/5)(3/5) = 4 1 P(Ā) = 5 5 Let X be number of heads observed in the twenty flippings. (a) Find the probability P(X = 8). (b) Find its mean and standard deviation. Solution: (a) According to the binomial distribution formula, we have P(A) = P(X = 8) = 8 4 20−8 20 1 C8 ( ) ( ) 5 5 (b) Its mean is equal to µ = 20(1/5) = 4 = 0.022. 8 and its standard deviation is equal to r p 16 σ = 20(1/5)(4/5) = 5 Example: A factory employs 2 thousand workers, of whom 30% are Hispanic. If the 15 members of the union executive committee were chosen from the workers at random,let X be the number of Hispanics on the committee. Find P(X = 3). Solution: This is not a binomial model because it is drawing sample without replacement. 1400 C3600 × C12 . P(X = 3) = 2000 C15 When the number of trials n is very big, it is very difficult to calculate the probability P(X = k) = Cknpk (1 − p)n−k . However, we have following approximation method. Normal approximation to binomial distributions Let X denote the count of successes in the binomial experiment with n trials and success probability p. If np ≥ 10 and n(1 − p) ≥ 10, thenp the distribution of X is approximately Normal, N (np, np(1 − p)). More precisely, we have following formula. For given constant integers a, b, use following formulae to find 9 approximate probabilities: (I) (II) a − np P(X ≥ a) = P(Z ≥ √ ) npq b − np P(X ≤ b) = P(Z ≤ √ ) npq Example: Consider the binomial experiment with n = 10, 000 and the probability of success p = 0.3. Let X be the number of successes in the 10, 000 trials. Find the approximate probability P(X ≤ 3100). Solution: (1) np = 10, 000(0.3) = 3, 000 > 10 and nq = 10, 000(0.7) = 7, 000 > 10. p √ (2) µ = np = 3, 000, σ = npq = 10, 000(0.3)(0.7) = 45.82 (3) To find the probability P(X ≤ 3, 100), according to (II), we have b − np ) P(X ≤ b) = P(Z ≤ √ npq 3, 100 − 3, 000 ) 45.82 = P(Z ≤ 2.19) = 0.9857 P(X ≤ 3, 100) = P(Z ≤ Example: Suppose that early statewide election returns indicate totals of 33, 00 votes for candidate A versus 27, 00 for candidate B, and that these early returns can be regarded as a random sample selected from the population of all 10, 000, 000 eligible voters in the state. 10 Let X be the number of votes for A. (a) If the statewide vote will be split 50 − 50, find the expected number of votes for A in the sample of 60, 000 early returns. (b) Find the standard deviation of X. (c) Find the probability that (X > 30, 100). Solution: (a) µ = np = 60, 000 × 0.5 = 30, 000. √ √ (b) σ = npq = 15, 000 = 122.47449 P 60,000 (c) P(X > 30, 0100) = 1 − 30,100 (0.5)i(0.5)60,000−i i=0 Ci However, we can find an approximate probability as follows: P(X > 30, 100) = 1 − P(X ≤ 30, 100) 30, 100 − 30, 000 = 1 − P(Z ≤ ) 122.47 = 1 − P(Z ≤ 0.8165) = 0.208