A Very Nice Practice Midterm Exam A representative sampling of the first half of prob and stat - Complete with follow-on questions to enhance the learning process Need to know in general sample spaces and random events counting methods Probabilities, means and variances, modes and medians joint distributions Probabilities, means and variances continuous random variables Combinations, permutations, fundamental counting rule condition probabilities, total prob rule, & Bayes theorem discrete random variables Unions, intersections, and complements covariance and correlation marginal and conditional distributions Descriptive statistics Question 1 The chief executive officer (CEO) of the Combinatorial Company, Mr. Hi N. Mitey, will be visiting 3 of his 5 plants this week. The plant manager of the Homestead Plant is betting that the CEO will not visit him this week and is therefore delaying preparations for his visit. If Hi N. Mitely selects his itinerary randomly, what is the probability that the Homestead Plant will not be visited this week? 4 1 favorable 3 0 4 Prob total 10 5 3 order of visits is not important Alternate approach favorable 4 3 2 2 Prob total 5 4 3 5 Follow-up to Question 1 The chief executive officer (CEO) of the Combinatorial Company decides not to visit the plants. Instead he will send his 3 vicepresidents. Each VP selects a plant at random to visit. What is the probability that the Homestead Plant is not selected? favorable 43 64 Prob 3 .512 total 5 125 Bonus Question #1 Faculty advisors are assigned randomly to new students entering the Engineering Management program from among five full-time faculty members. Three new students are admitted on a particular day. The office staff member, Dizzy Dunce, assigns (randomly) each student a faculty advisor. (a) What is the probability that each student is assigned to a different advisor? (b) What is the probability that all three students are assigned to one faculty member? (a) (5)(4)(3) / 53 = .48 (b) 5 / 53 = 1/25 = .04 (5 ways to select the same faculty member) Question 2 Eighty percent of all students taking probability and statistics pass the course while seventy-five percent of those taking operations research pass the course. However, of those that have passed probability and statistics, 90 percent will pass operations research. What fraction of students passes both courses? What fraction will pass at least one course? Let A = the event, student passes prob/stat B = the event, student passes operations research Given: P(A) = .80; P(B) = .75; P(B|A) = .90 Required: P(A B) = ? and P(A B) = ? P(A B) = P(B|A) P(A) = (.90) (.80) = .72 P(A B) = P(A) + P(B) - P(A B) = .80 + .75 - .72 = .83 Question 3 The Rockweed Aircraft Corporation is redesigning the instrument panel for its new F-222 steam propelled fighter. After interviewing a large number of pilots, they decided to replace certain gauges with idiot (warning) lights. One such light warns of engine trouble. There is a probability of .01 that there will be engine trouble for a single mission of this aircraft. Given there is engine (boiler) trouble, there is a .99 probability that the idiot light turns on. If there is no engine trouble during the mission, there is a .98 probability that the idiot light will not turn on. What is the probability that the engine light turns on during a mission. If the engine light turns on during a mission, what is the probability of engine trouble? Question 3 (continued) Let T = the event, engine trouble O = the event, engine light turns on Given: P(T) = .01; P(O|T) = .99; P(Oc|Tc) = .98 Required: P(O) and P(T|O) The Total Probability Rule (TPR): P(O) = P(O|T) P(T) + P(O|Tc) P(Tc) P(O) = .99 (.01) + .02 (.99) = .0297 Then Bayes: P(T|O) = P(O|T) P(T) / P(O) = .99 (.01) / .0297 = .333 Question 4 The table summarizes the GPA and statistics aptitude test results of students admitted to the grad engineering programs. Based upon these historical numbers, determine the relative frequency probability that an individual applying for a grad engineering program: (a) will score low on the aptitude test and have a GPA above 3.5 (b) will score high on the aptitude test if individual has a GPA between 3.0 and 3.5 (c) Show that test scores and GPA are not independent events Entering Grade Point Average (GPA) Aptitude High test results Medium Low Below 3.0 3.0 – 3.5 Above 3.5 0 17 13 5 40 5 15 3 2 Question 4 (a) will score low on the aptitude test and have a GPA above 3.5 (b) will score high on the aptitude test if individual has a GPA between 3.0 and 3.5 (c) Show that test scores and GPA are not independent events (a) P(score low and GPA > 3.5) = 2/100 (b) P(score high | 3<GPA<3.5) = 17/60 (c) P(High|GPA<3) = 0 P(High) = 30/100 Entering Grade Point Average (GPA) Below 3.0 3.0 – 3.5 Above 3.5 0 17 13 30 5 40 5 50 Low 15 3 2 20 total 20 60 20 Aptitude High test results Medium Question 5 The probability that a missile fired at a terrorist target will destroy the target is .6. Assuming independence, what is the minimum number of missiles to be launched against the target in order to have at least a 90 percent chance of destroying the target? Let X = the discrete random variable, f ( x) .6 .4 x 1 , F ( x) 1 .4 x the number of missiles fired until 1 Pr{ X 1} F (1) 1 .4 .60 the target is destroyed. X ~ Geo(.6) Pr{ X 2} F (2) 1 .4 .84 2 Pr{ X 3} F (3) 1 .4 .936 3 Pr{ X 4} F (4) 1 .4 .9744 4 Pr{ X 5} F (5) 1 .4 .98976 5 Follow-on Question 5 The probability that a missile fired at a terrorist target will destroy the target is .6. Assuming independence, what is the probability of destroying the target with the 3rd missile? What is the expected number of missiles needed to destroy the target? Let X = the discrete random variable, the number of missiles fired until the target is destroyed. X ~ Geo(.6) f ( x) .6 .4 x 1 , F ( x) 1 .4 Pr{ X 3} f (3) .6 .4 1 E[ x] 1.667 .6 31 x .096 Bonus Question #2 The Catastrophic Construction Company provides emergency repair of facilities damaged during natural disasters such as hurricanes and tornados. They respond primary to requests from FEMA (Federal Emergency Management Agency). The number of such requests per year is a random variable best described by the following probability mass function (PMF): 2 x 4 f ( x) 31 , x 0,1, 2,3, 4,5 E[X] = 0 (16/31) + 1(9/31) + 2(4/31) + 3(1/31) + 4(0/31) + 5(1/31) = 25/31 = .806 E[X2] = 0 (16/31) + 1(9/31) + 4(4/31) + 9(1/31) + 16(0/31) + 25(1/31) = 1.903 V[X] = 1.903 - .8062 = 1.253, = 1.12 note correction Question 6 The Loose Screw Company manufactures nuts and bolts for sale to wholesale distributors. The delivery time (lead-time) for orders in days is given by the following Probability Density Function (PDF): f(x) = x/300 , 5 x 25 Find the probability that delivery will be within one standard deviation of the mean. Question 6 x x z z x 2 25 f ( x) , 5 x 25; F ( x) dz 300 300 600 5 600 5 x 25 x x 15625 125 dx 17.222 5 300 900 900 5 25 E[ X ] 2 3 25 x x 390625 625 dx 325 5 300 1200 1200 5 25 E[ X 2 ] 2 3 4 2 325 17.2222 28.402, 5.3294 Pr{17.222 5.3294 x 17.222 5.3294} Pr{11.89 x 22.55} 22.552 25 11.892 25 F (22.55) F (11.89) .80584 .19395 .612 600 600 f(x) = x/300 , 5 x 25 The Geometry of Question 6 25/300 f(x) 20/300 5/300 x 5 25 Area = (20)(5/300) + (1/2)(20)(20/300) =1 F ( x) 5 1 x 5 x 5 x 5 300 2 300 (5)(2) x 5 x 5 x 5 x 2 25 600 600 x Question 7 Given the following joint probability mass function, find the correlation between X and Y. X/Y 0 1 totals E[Y] = E[XY] = covar = y 0 1 2 f(y) .3 .4 .3 E[Y] = 0 (.3) + 1(.4) + 2(.3) = 1.0 V[Y] = 0 (.3) + 1(.4) + 4(.3) - 12 = .6 X/Y 0 1 2 0 .1 .3 .2 1 .2 .1 .1 0 0.1 0.2 1 0.3 0.1 2 totals 0.2 0.6 0.1 0.4 0.3 0.4 0.3 1 E[Y^2] = 1.6 0.3 -0.1 Cor[XY] = -0.263523 E[X] E[X^2} 0.4 0.4 V[X] = 0.24 V[Y] = 0.6 Question 8 Ms. Ima Borne Loser makes correct decisions 10 percent of the time. How many (independent) decisions must she make to have at least a 90 percent chance of making at least one correct decision? Hint: let Ei = the event, the ith decision is correct. Given: P( Ei ) .10, therefore P( Eic ) .9 Required : P( E1 E2 ... En ) .9, find n P( E1 E2 ... En ) 1 P( E1c E2c ... Enc ) 1 P( E1c ) P( Enc ) 1 .9n .9 .9n .1 ln .1 n ln .9 ln .1 or n 21.85; n 22 ln .9 Question 8 – alternate approach Ms. Ima Borne Loser makes correct decisions 10 percent of the time. How many (independent) decisions must she make to have at least a 90 percent chance of making at least one correct decision? Let X = a discrete RV, the number of correct decisions in n trials X B(n,.1) Pr X 1 1 Pr X 0 1 (1 .1) n 1 .9n .9 n 10 15 20 25 Pr{X 1} .6513 .7941 .8784 .9282 n 21 22 Pr{X 1} .8906 .9015 Question 9 Thirty percent of the Fly-By-Nite Airline Company’s flights are delayed. If Mr. I. N. Hurrie is scheduled on 5 flights as he travels to each of the company’s 4 technical centers, what is the probability of no more than one flight being delayed? Assume Independence. Let X = a discrete RV, the number of delayed flights among the 5 X ~ B(5, .3) Pr{X 1} = f(0) + f(1) = F(1) = .528 (from the Prob Calculator) What is the expected number of delayed flights? E[X] = (5)(.3) = 1.5 Question 10 Given the following CDF, find the mean of the probability distribution. x2 F ( x) , 0 x 10 100 dF ( x) x f ( x) dx 50 10 3 10 x x 1000 E[ X ] x dx 6.67 50 150 0 150 0 Follow-on Question 10 Given the following CDF, find the variance of the probability distribution. x2 F ( x) , 0 x 10 100 10 4 10 x x 10000 E[ X ] x dx 50 50 200 0 200 0 2 2 V [ X ] 50 6.6667 5.55511; 2.3569 2 Follow-on to the Follow-on of Question 10 Given the following CDF, find the mean and variance of the probability distribution. Isn’t this just a right triangular distribution with b = 10? x2 x F ( x) , 0 x 10; f ( x) 100 50 2x f ( x) 2 0 x b b x2 F ( x) 2 2 Var[X] = 2.357 b from Prob Calc 2 Beta ( =2 =1) E[ X ] b 3 Question 11 Professor I. Do Little’s research (NSF Grant) on the migration pattern of the coastal plain swamp sparrow has determined that their migration (flying) time is normally distributed with a mean of 23 days and a variance of 36 days. Compute the probability that a particular swamp sparrow will migrate between 30 to 36 days. X n(23,36) Pr{30 X 35} F (35) F (30) 35 23 30 23 Pr z Pr z Pr z 2 Pr z 1.17 6 6 .9772 .8783 .0989 Follow-on Question 11 Professor I. Do Little’s research (NSF Grant) on the migration pattern of the coastal plain swamp sparrow has determined that their migration (flying) time is normally distributed with a mean of 23 days and a variance of 36 days. 90 percent of the sparrow population will complete their migration in how many days? X n(23,36) Pr{ X x} .90 x 23 Pr z .90; z 1.2816 6 x z 23 1.2816 6 30.6896 days Question 12 Ted E. Bare, an engineering student, has observed that the number of times the campus police check the parking lot has a Poisson distribution with a mean of once every 4 hours. If Ted E. is parked illegally during a 3-hour evening class, what is the probability he will get ticketed? Let X = a discrete RV, the number of times in a 3-hour period, the Campus police will check the parking lot. X Pois ( 3 x .25) Pr X 1 1 f (0) 1 e .75 .75 0! 0 1 e .75 .5276 Follow-on to Question 12 At break time, one and half hours into the class, Ted checks and sees that he did not receive a ticket as yet. What is the probability that he will get ticketed by the time class is out? Let X = a discrete RV, the number of times in a 1.5-hour period, the Campus police will check the parking lot. X Pois ( 1.5 x .25) Pr X 0 e .375 .375 0! Prob get ticket = 1 - .6873 = .3127 0 e .375 .6873 Also, let Y = a continuous RV, the time to the next arrival of the campus police Y ~ Exp( = .25) and F(y) = 1 – e-.25y Therefore Pr{Y<1.5} = 1 – F(1.5) = 1 - e-.25(1.5) = .3127 Question 13 Dawn E. Brook is an undergraduate student who works parttime in the Engineering Management office. Among her duties is to make coffee for the faculty and staff. The time it takes the old coffee pot to make coffee is a random variable having a Weibull distribution with = 2 and = 20 minutes. If the coffee pot perks for less than 16 minutes, the coffee will be too weak and if the coffee pot perks for more than 23 minutes, the coffee will be too strong. What is the probability that Dawn E. Brook will brew a pot of coffee to the faculty’s liking? X Weib(2, 20) Pr 16 X 23 F (23) F (16) .73353 .47271 .261 Alternate Question 13 Dawn E. Brook is an undergraduate student who works parttime in to the Engineering Management office. Among her duties is to make coffee for the faculty and staff. The time it takes the old coffee pot to make coffee is a random variable having the following cumulative distribution (CDF): F ( x) 1 e x 20 2 , x0 If the coffee pot perks for less than 16 minutes, the coffee will be too weak and if the coffee pot perks for more than 23 minutes, the coffee will be too strong. What is the probability that Dawn E. Brook will brew a pot of coffee to the faculty’s liking? F (23) F (16) .73353 .47271 .261 What is the mean brewing time? X Weib(2, 20) E[ X ] 17.72 min . from Prob Calc Question 14 Laye Z. Jones has been given three tasks to complete within the next 8 hours. The time for Laye Z. to complete each task based upon past performance is normally distributed with the following parameters: What is the probability that Laye Z. will complete all 3 task on time. Y X1 X 2 X 3 Y Task Mean Variance 1 3.4 hr. 2.4 2 4.8 hr. 3.8 3 2.5 hr. 1.8 Totals 10.7 8.0 n(10.7,8) 8 10.7 Pr Y 8 Pr z Pr z .9546 .1699 8 100 E[ X ] 0 3x 2 x 6 10 100 4 3 3 x 3 dx 6 x dx 6 10 4 x 10 0 Question 15 100 0 300 75 days 4 5000 66.67 75 75 The time it takes to complete an engineering design project in days is a random variable having the following probability density function (PDF): 3x 2 for 0 x 100 days f ( x) 106 0 otherwise The profit resulting from completion of the project is given by $5,000 / X. Find the expected profit. 100 E[Pr ofit ] 0 15, 000 x 2 x 106 2 100 0 100 2 15000 5000 3x x dx 6 dx 6 10 0 x 10 15, 000 $75 2 x 100 Bonus Question #3 Given the following very fine probability density function (PDF) where the random variable X is Professor Domkoff ’s driving time in hours to school, find the (a) mean and the (b) probability that the driving time is no more than 90 minutes. 2 f ( x) 2 , 1 x 2 x 2 2x 2 2 dx 2 ln x 1 2 ln 2 2 ln1 1.3863 hr x 1 x 2 2 2 2 2 F ( x) 2 dz 2 z z 1 x 1 x 1 x F (1.5) 2 2 .667 1.5 Problem 16 – Descriptive Stats & point estimation The following random sample was obtained by measuring the time in (working) hours to complete a particular construction job. Treating the data as continuous, answer the following questions: (a) Find a sample estimate for the population mean (b) Find a sample estimate for the population variance (c) Is the data skewed left, right, or almost symmetrical? (d) Find the sample interquartile range. (e) Find the sample median. (f) Based only on a histogram, which of the following distributions are not likely models for the population distribution? Normal Weibull Rectangular Exponential Triangular A Truly Great Set of Data 83.8 64.2 89.8 82.3 90.4 63.3 40.4 104 108 96.6 65.4 98.1 46.8 86.9 71.8 56.2 72.1 73.7 77.2 113 135 56.4 99.8 64.6 95.7 85.3 75.8 88.5 71.7 72 99.7 49.1 98.9 85.2 110 68.4 123 58.6 74.1 67.9 66.5 44.5 55.6 136 91.7 Task Time in Hours 30.9 76.7 36.8 48 71.5 Problem 16 sample size mean variance std dev median 1st quartile 3rd quartile interquatile range Mimimum Maximum Range Skewness Kurtosis 50 78.420 569.696 23.868 74.95 64.3 95.7 31.4 30.9 135.7 104.8 0.3234 -0.0294 sturges rule = interval width = 6.606601 17.46667 6 17 18 16 14 12 10 8 6 4 2 0 47 64 81 98 115 132 149 From the histogram rule out the rectangular and exponential - also compare mean and std. dev. some skewness to the right Keys to successful completion of the midterm Work every problem if unable to complete – submit a partial or intermediate answer state any assumptions, identify type of problem anything submitted correctly about problem will earn some credit Apply common sense to your answers probabilities cannot exceed one, variances are positive does the magnitude of the number make sense Use the computer and Prob Calculator no credit for simply restating the problem saves time, less chance of an error Start by defining events or RV’s write down what’s given follow up with what is required Key Things to know Recognize problem type Recognize types of random events random events versus random variables dependent versus independent events conditional versus unconditional probability total probability problem (weighted average) Bayes problem Recognize probability distributions theoretical density and distribution functions i.e. Poisson or uniform PMF; exponential or Weibull PDF distinguish between binomial and geometric joint versus marginal versus condition distributions The Important Distributions The discrete ones Uniform Binomial Geometric Poisson The continuous ones Rectangular Normal Exponential Weibull Triangular Left triangular Right triangular Key things to expect A combinatorial problem or an a priori probability A random event problem find mean, median, variance, probabilities, median A joint distribution – most likely discrete conditional or Bayes independent A binomial, geometric or both A normal distribution problem A general distribution # favorable /total use of counting methods find marginal, conditional – probability, mean, variance, correlation A problem requiring integration find a mean or probability