Probability Assignment SOLUTION) I) We are given the cumulative distribution function (CDF) F(x) of the discrete random variable X. We can determine the probability mass function (PMF) f(x) from the CDF as follows: 𝑓(𝑥) = 𝐹(𝑥) − 𝐹(𝑥 − 1) Applying this, we get: Probability Assignment 𝑓(0) = 𝐹(0) − 𝐹(−1) = 0 − 0 = 0 𝑓(1) = 𝐹(1) − 𝐹(0) = 0.1 − 0 = 0.1 𝑓(2) = 𝐹(2) − 𝐹(1) = 0.7 − 0.1 = 0.6 𝑓(3) = 𝐹(3) − 𝐹(2) = 0.7 − 0.7 = 0 𝑓(4) = 𝐹(4) − 𝐹(3) = 0.7 − 0.7 = 0 𝑓(5) = 𝐹(5) − 𝐹(4) = 0.7 − 0.7 = 0 𝑓(6) = 𝐹(6) − 𝐹(5) = 1 − 0.7 = 0.3 We therefore get the following PMF: 0.1 𝑥 = 1, 𝑓(𝑥) = 𝑃(𝑋 = 𝑥) = 0.6 𝑥 = 2, {0.3 𝑥 = 6, and 0 otherwise. We can now easily determine each of the required probabilities: a) 𝑃(𝑋 ≤ 3.5) = 𝑃(𝑋 ≤ 3) = 𝑓(1) + 𝑓(2) = 0.1 + 0.6 = 𝟎. 𝟕 b) 𝑃(𝑋 > 5) = 𝑓(6) = 𝟎. 𝟑 c) 𝑃(𝑋 ≤ 5.2) = 𝑃(𝑋 ≤ 5) = 𝑓(1) + 𝑓(2) = 0.1 + 0.6 = 𝟎. 𝟕 d) 𝑃(𝑋 > 7) = 𝟎 e) 𝑃(1 ≤ 𝑋 ≤ 2) = 𝑓(1) + 𝑓(2) = 0.1 + 0.6 = 𝟎. 𝟕 f) 𝑃(𝑋 < 2) = 𝑓(1) = 𝟎. 𝟏 g) 𝑃(𝑋 = 2) = 𝑓(2) = 𝟎. 𝟔 h) 𝑃(𝑋 ≤ 2) = 𝑓(1) + 𝑓(2) = 0.1 + 0.6 = 𝟎. 𝟕 Probability Assignment II) The mean 𝜇 and variance σ2 are given by: 𝜇 = 𝐸[𝑋] = ∑ 𝑥𝑓(𝑥) 𝑥 𝜎 2 = ∑(𝑥 − 𝜇)2 𝑓(𝑥) 𝑥 Therefore, the mean is: 𝜇 = 1(0.1) + 2(0.6) + 6(0.3) = 𝟑. 𝟏 The variance is: 𝜎 2 = (1 − 3.1)2 (0.1) + (2 − 3.1)2 (0.6) + (6 − 3.1)2 (0.3) = 𝟑. 𝟔𝟗 Probability Assignment SOLUTION) a) The distribution that describes the random variable X representing the number of test repetitions until the student achieves the University requirement is the geometric distribution, with the probability of success on each independent trial being p = 0.3. 𝑋 ~ 𝑔(𝑥; 0.3) The PMF of X is: 𝑓(𝑥) = 0.3(0.7)𝑥−1 𝑥 = 1, 2, 3, … Probability Assignment b) The average number of trials by any student to achieve the requirements is the mean 𝜇 of the random variable X. The mean of any geometric random variable is the reciprocal of p, therefore: 𝜇= 1 1 = = 𝟑. 𝟑𝟑 𝑝 0.3 c) The probability that a student takes the test 4 times until the requirement is achieved is found by: 𝑃(𝑋 = 4) = 𝑔(4; 0.3) = 0.3(0.7)4−1 = 𝟎. 𝟏𝟎𝟐𝟗 d) The probability that a student achieves the requirement at the 6th trial given that he did not achieve it in the first and second trails is: 𝑃(𝑋 = 6|𝑋 > 2) = = 𝑃(𝑋 = 6) ∩ 𝑃(𝑋 > 2) 𝑃(𝑋 = 6) = 𝑃(𝑋 > 2) 𝑃(𝑋 > 2) 𝑃(𝑋 = 6) 𝑃(𝑋 = 6) = 1 − 𝑃(𝑋 ≤ 2) 1 − 𝑃(𝑋 = 1) − 𝑃(𝑋 = 2) 0.3(0.7)6−1 0.3(0.7)5 0.3(0.7)5 = = = 1 − 0.3(0.7)1−1 − 0.3(0.7)2−1 1 − 0.3 − 0.3(0.7) 0.7(1 − 0.3) 0.3(0.7)5 = = 0.3(0.7)3 = 𝟎. 𝟏𝟎𝟐𝟗 2 0.7 Probability Assignment e) We can see that the answers to part (c) and (d) are the same. This is due to the memoryless property of the geometric distribution. In part (d), we are given that the student did not meet the requirements in the first and second trials. He will therefore need 4 more trials to achieve the requirements at the 6th trial. Because of the memoryless property of the geometric distribution, the probability of this happening is the same as that of achieving the requirements at the 4th trial, which is the case in part (c). f) The random variable Y representing the amount the student needs to pay in dollars is given by: 𝑌 = 100 + 20𝑋 The expectation of Y is: 𝐸[𝑌] = 𝐸[100 + 20𝑋] = 𝐸[100] + 𝐸[20𝑋] = 100 + 20𝐸[𝑋] 100 + 20 ( 1 ) = 𝟏𝟔𝟔. 𝟔𝟕 0.3 The variance of Y is: 𝜎 2 [𝑌] = 𝜎 2 [100 + 20𝑋] = 202 𝜎 2 [𝑋] = 400 ( 1 − 0.3 ) = 𝟑𝟏𝟏𝟏. 𝟏𝟏 0.32 Probability Assignment SOLUTION) a) Let X denote the binomial random variable representing the number of free lifts the student will get: 𝑋 ~ 𝑏(𝑥; 𝑛, 𝑝) where n is the number of trials and p is the probability of success on a trial. 𝑛 𝑓(𝑥) = ( ) 𝑝 𝑥 𝑞 𝑛−𝑥 𝑥 𝑥 = 0, 1, 2, … , 𝑛 where q = 1 – p. We are told that the probability of a free lift is 0.4, hence p = 0.4. The probability that the student will get a lift only twice in a week of five days (n = 5) is: 5 𝑃(𝑋 = 2) = 𝑏(2; 5, 0.4) = ( ) (0.4)2 (1 − 0.4)5−2 = 𝟎. 𝟑𝟒𝟓𝟔 2 Probability Assignment b) Let Y denote the geometric random variable representing the day on which the student gets his/her first free lift: 𝑌 ~ 𝑔(𝑥; 𝑝) where p is the probability of success on a trial. 𝑓(𝑥) = 𝑝𝑞 𝑥−1 𝑥 = 0, 1, 2, … where q = 1 – p. As earlier, p = 0.4. The probability that the student will get his/her first lift on the third day is therefore: 𝑃(𝑌 = 3) = 0.4(1 − 0.4)3−1 = 0.4(0.6)2 = 𝟎. 𝟏𝟒𝟒 c) The number of times the student needs to get a taxi is a binomial random variable. Since the student needs to get a taxi if he/she does not get a lift, the probability of getting a taxi is 1 – 0.4 = 0.6. The expected value of a binomial random variable is 𝜇 = 𝑛𝑝. Therefore, in 20 days, we expect that the student will need to get a taxi 0.6(20) = 12 times. We are told that the student needs to pay 20 QAR for the taxi, so the expected amount of money to be paid in 20 days is: 12(20) = 𝟐𝟒𝟎 𝐐𝐀𝐑 Probability Assignment SOLUTION) Let X be the random variable representing the number of calls coming per minute into a hotel reservation. We are told that this random variable would follow a Poisson distribution. Therefore, X has the following PDF: 𝑒 −𝜆𝑡 (𝜆𝑡)𝑥 𝑓(𝑥) = 𝑥! 𝑥 = 0, 1, 2, 3, … where 𝜆 is the average number of calls coming per minute. a) We are told that the mean number of calls per minute is 3, which gives us the value of 𝜆. As for t, this would simply be 1 since it is a 1-minute period. The probability that no calls come in a given 1-minute period is: 𝑒 −3 (3)0 𝑃(𝑋 = 0) = = 𝑒 −3 = 𝟎. 𝟎𝟒𝟗𝟖 0! Probability Assignment b) In this part, the period is of length two minutes and so we have to adjust the value of t to 2. We would thus have 𝜆𝑡 = 3(2) = 6. The probability that at least two calls will arrive in a given two-minute period is: 1 𝑒 −6 (6)𝑥 𝑃(𝑋 ≥ 2) = 1 − 𝑃(𝑋 < 2) = 1 − ∑ = 𝟎. 𝟗𝟖𝟐𝟔 𝑥! 𝑥=0 c) In this part, the period is of length five minutes and so we have to adjust the value of t to 5. We would thus have 𝜆𝑡 = 3(5) = 15. The probability that at most two calls will arrive in a given five-minute period is: 2 𝑒 −15 (15)𝑥 𝑃(𝑋 ≤ 2) = ∑ = 𝟑. 𝟗𝟑𝟎𝟖 × 𝟏𝟎−𝟓 𝑥! 𝑥=0 Probability Assignment SOLUTION) The probability density function (PDF) can be obtained from the cumulative distribution function (CDF) by taking its derivative: 𝑓(𝑥) = 𝑑𝐹(𝑥) 𝑑𝑥 Taking the derivative for each interval, we end up with the following PDF: 𝑓(𝑥) = 0.2 0≤𝑥<4 0.05 4≤𝑥<8 {0 otherwise We now find the mean 𝜇: ∞ 4 8 𝜇 = 𝐸[𝑋] = ∫ 𝑥𝑓(𝑥) 𝑑𝑥 = ∫ 0.2𝑥 𝑑𝑥 + ∫ 0.05𝑥 𝑑𝑥 −∞ 4 8 = 0.1𝑥 | + 0.025𝑥 | = 0.1(16) + 0.025(64 − 16) = 𝟐. 𝟖 0 4 2 4 0 2 Probability Assignment The variance 𝜎 2 is defined as: 𝜎 2 = 𝐸[𝑋 2 ] − (𝐸[𝑋])2 = 𝐸[𝑋 2 ] − 𝜇 2 We first find 𝐸[𝑋 2 ]: ∞ 𝐸[𝑋 2] 4 2 8 = ∫ 𝑥 𝑓(𝑥) 𝑑𝑥 = ∫ 0.2𝑥 𝑑𝑥 + ∫ 0.05𝑥 2 𝑑𝑥 −∞ 2 0 4 1 3 4 1 3 8 1 3 1 3 176 (4 ) + (8 − 43 ) = = 𝑥 | + 𝑥 | = 15 60 15 60 15 0 4 ∴ 𝜎2 = 176 𝟐𝟗𝟐 − (2.8)2 = ≅ 𝟑. 𝟖𝟗𝟑 15 𝟕𝟓 Probability Assignment SOLUTION) The uniformly distributed random variable X representing the weight of motors delivered by the shipping company is given as: 0.1 𝑓(𝑥) = { 0 1200 < 𝑥 < 1210 otherwise a) The mean motor weight is: ∞ 1210 𝜇 = 𝐸[𝑋] = ∫ 𝑥𝑓(𝑥) 𝑑𝑥 = ∫ −∞ 2 1210 0.1𝑥 𝑑𝑥 = 0.05𝑥 | 1200 1200 ∴ 𝜇 = 0.05(1210)2 − 0.05(1200)2 = 𝟏𝟐𝟎𝟓 Alternatively, we could have used the fact that the mean of any uniformly distributed random variable over the interval (A, B) is simply the average of A and B. In our case, A and B are 1200 and 1210 respectively. Their average is 1205 which agrees with the result obtained by integration. Probability Assignment The standard deviation is the positive root of the variance. Therefore, to obtain the standard deviation of X, we first find its variance: 𝜎 2 = 𝐸[𝑋 2 ] − (𝐸[𝑋])2 = 𝐸[𝑋 2 ] − 𝜇 2 ∞ 𝐸[𝑋 2] 1210 2 = ∫ 𝑥 𝑓(𝑥) 𝑑𝑥 = ∫ −∞ = 1200 1 3 1210 0.1𝑥 𝑑𝑥 = 𝑥 | 30 1200 2 1 4356100 (12103 − 12003 ) = 30 3 ∴ 𝜎2 = 4356100 25 − (1205)2 = 3 3 Alternatively, we could have used the fact that the variance of any uniformly distributed random variable over the interval (A, B) is: (𝐵 − 𝐴)2 (1210 − 1200)2 25 𝜎 = = = 12 12 3 2 which agrees with the result found by integration. In any case, the standard deviation can be readily obtained from the variance by taking its positive square root: 𝜎 = √𝜎 2 = √ 25 𝟓√𝟑 = ≅ 𝟐. 𝟖𝟖𝟕 3 𝟑 Probability Assignment b) The proportion of motors within specifications is: 1205 𝑃(1195 < 𝑋 < 1205) = ∫ 1195 1205 𝑓(𝑥) 𝑑𝑥 = ∫ 0.1 𝑑𝑥 1200 1205 = 0.1𝑥 | = 0.1(1205 − 1200) = 𝟎. 𝟓 1200 Therefore, half of the motors will be within the weight specifications. Probability Assignment SOLUTION) Let the random variable T denote the reaction time. We are told that this random variable follows a normal distribution with a mean of 0.5 seconds and a standard deviation of 0.05 seconds. a) The required probability is P(T > 0.5). Since T is a normal random variable, we need to convert it to the standard normal random variable Z in order to use the table providing the areas under the normal curve: 𝑃(𝑇 > 0.5) = 𝑃 (𝑍 > 0.5 − 0.5 ) = 𝑃(𝑍 > 0) = 𝟎. 𝟓 0.05 We could have arrived at this result straight away because 0.5 seconds is the mean; the probability that any normal random variable exceeds its mean is always 50%. Probability Assignment b) The required probability is P(0.496 < T < 0.502). Again, we convert to the standard normal random variable Z in order to use the table values: 𝑃(0.496 < 𝑇 < 0.502) = 𝑃 ( 0.496 − 0.5 0.502 − 0.5 <𝑍< ) 0.05 0.05 = 𝑃(−0.08 < 𝑍 < 0.04) = 𝑃(𝑍 < 0.04) − 𝑃(𝑍 < −0.08) = 0.5160 − 0.4681 ∴ 𝑃(0.496 < 𝑇 < 0.502) = 𝟎. 𝟎𝟒𝟕𝟗 c) Our goal here is to find the value a such that: 𝑃(𝑇 > 𝑎) = 0.90 This is equivalent to: 𝑃(𝑇 < 𝑎) = 0.10 Converting to the standard normal random variable Z and using the table, we get: 𝑃 (𝑍 < 𝑎 − 0.5 𝑎 − 0.5 ) = 0.10 → = −1.28 0.05 0.05 ∴ 𝑎 = −1.28(0.05) + 0.5 = 𝟎. 𝟒𝟑𝟔 The reaction time that is exceeded 90% of the time is therefore 0.436 seconds. Probability Assignment SOLUTION) Let X denote a hypergeometric random variable representing the number of pages in error within the 100 pages sampled randomly from the lot of 1000 pages: 𝑋 ~ ℎ(𝑥; 𝑁, 𝑛, 𝑘) 𝑓(𝑥) = (𝑘𝑥 )(𝑁−𝑘 ) 𝑛−𝑥 (𝑁 ) 𝑛 Here, n is the size of the random sample of pages selected, N is the total number of pages, of which k are the pages in error and N – k are the pages not in error. The probability that at least 30 of the pages in error is in the randomly selected sample is P(X ≥ 30). In this problem, we have N = 1000 and n = 100. In such cases when n is small relative to N, the binomial distribution can be used to approximate the hypergeometric distribution. We are told that the printer produces an error on 50 of 1000 pages, so we take p, the binomial parameter which is the probability that a page is in error to be: 𝑝= 𝑘 50 = = 0.05 𝑁 1000 Probability Assignment We can also use another approximation; because n is large, the binomial distribution can be approximated well by the normal distribution. If X is a binomial random variable whose mean is 𝜇 = 𝑛𝑝 and variance is 𝜎 2 = 𝑛𝑝𝑞 = 𝑛𝑝(1 − 𝑝), then the following standard normal random variable Z can be used to approximate the binomial distribution: 𝑍= 𝑋 − 𝑛𝑝 √𝑛𝑝(1 − 𝑝) The required probability can therefore be approximated as: 𝑃(𝑋 ≥ 30) = 1 − 𝑃(𝑋 < 30) = 1 − ℎ(< 30; 1000, 100, 50) ≈ 1 − 𝑏(< 30, 100, 0.05) ≈ 1 − 𝑃 (𝑍 < 29.5 − 100(0.05) ) = 1 − 𝑃(𝑍 < 11.24) ≅ 1 − 1 = 𝟎 √100(0.05)(1 − 0.05) Note that a continuity correction of 0.5 was used in the conversion from the discrete binomial distribution to the continuous normal distribution (30 to 29.5). Note as well that in reality, the probability is extremely small but not exactly 0; the calculations merely lack the precision required for the exact extremely small value. This makes sense because if only 50 pages are in error out of the total 1000 pages, then it is extremely unlikely for there to be 30 or more pages in error within a sample of just 100 pages. Probability Assignment If the question was instead to find the probability that at least one of the pages in error is in the sample, we would have proceeded as follows: 𝑃(𝑋 ≥ 1) = 1 − 𝑃(𝑋 < 1) = 1 − 𝑃(𝑋 = 0) = 1 − ℎ(0; 1000, 100, 50) ≈ 1 − 𝑏(0, 100, 0.05) ≈ 1 − 𝑃 (𝑍 < 0.5 − 100(0.05) √100(0.05)(1 − 0.05) ) = 1 − 𝑃(𝑍 < −2.06) ≅ 1 − 0.0197 = 𝟎. 𝟗𝟖𝟎𝟑 If we were to compute the probability using the hypergeometric distribution without any approximations, the result would be: 𝑃(𝑋 ≥ 1) = 1 − 𝑃(𝑋 < 1) = 1 − 𝑃(𝑋 = 0) 1− (50 )(950 ) 0 100 (1000 ) 100 = 1 − 0.00448 = 𝟎. 𝟗𝟗𝟓𝟓𝟐 We can therefore see how the approximation did a fairly reasonable job. Probability Assignment SOLUTION) We can let the random variable X denote the distance between major cracks in a highway. We are told that this random variable follows an exponential distribution with a mean of 10 km. We can therefore express the PDF of X as follows: 1 −𝑥 𝑒 10 10 𝑓(𝑥) = { 𝑥>0 o. w. 0 a) The probability that there are no major cracks in a 20 km stretch of the highway means the distance between major cracks is more than 20 km. The required probability is therefore: 𝑥 ∞ 1 −𝑥 −10 10 𝑃(𝑋 > 20) = ∫ 𝑓(𝑥) 𝑑𝑥 = ∫ 𝑒 𝑑𝑥 = −𝑒 | 20 20 10 20 ∞ = −𝑒 −∞ + ∞ 20 −10 𝑒 = 𝑒 −2 = 1 = 𝟎. 𝟏𝟑𝟓𝟑 𝑒2 Probability Assignment b) Using the relationship between the exponential probability distribution and the Poisson process, we can consider a random variable Y representing the number of major cracks in a 20 km stretch of the highway. From the mean of X, we can see that we expect an average of one major crack per 10 km. Therefore, in a 20 km stretch of the highway, we expect the mean number of major cracks to be two. We conclude that for the Poisson random variable Y, we have 𝜆𝑡 = 2. The PDF of Y would therefore be: 𝑒 −2 (2)𝑦 𝑓(𝑦) = 𝑦! 𝑦 = 0, 1, 2, … Using this, we can easily calculate the probability that there are two major cracks in a 20 km stretch: 𝑒 −2 (2)2 2 𝑃(𝑌 = 2) = = 2 = 𝟎. 𝟐𝟕𝟎𝟕 2! 𝑒 c) The fact that there were no major cracks in the first 10 km inspected has no impact on the likelihood of major cracks in the next 20 km. This is because the exponential distribution is memoryless. We therefore only have to find the probability that no major cracks occur in a 20 km stretch, which is essentially the same as part (a): 𝑃(𝑋 > 20) = 𝑒 −2 = 𝟎. 𝟏𝟑𝟓𝟑