Math 151 Homework 3 Solutions (Winter 2015) Problem 4.6 For k = 0, 1, 2, 3 if there are k heads then there must be 3 − k tails and in this case X = k − (3 − k) = 2k − 3. So X can get values −3, −1, 1, 3 and if H is the number of heads then 3 1 3 1 P (X = −3) = P (H = 0) = = 0 2 8 3 3 3 1 = P (X = −1) = P (H = 1) = 2 8 1 3 3 1 3 P (X = 1) = P (H = 2) = = 2 2 8 3 3 1 1 P (X = 3) = P (H = 3) = = 3 2 8 Problem 4.21 (a) E[X] will be bigger than E[Y ] because Y gets each of 40, 33, 25 and 50 with a constant probability whereas X takes each of these values with probability proportional to it i.e. big values with bigger probability. (b) E[X] = 40P (X = 40) + 33P (X = 33) + 25P (X = 25) + 50P (X = 50) 40 33 25 50 5814 = 40 · + 33 · + 25 · + 50 · = 148 148 148 148 148 = 39.28 and E[Y ] = 40P (Y = 40) + 33P (Y = 33) + 25P (Y = 25) + 50P (Y = 50) 1 1 1 1 = 40 · + 33 · + 25 · + 50 · 4 4 4 4 = 37 1 Problem 4.26 (a) Observe that to guess the values i we need to ask exactly i questions. So X = i with 1 for all 1 ≤ i ≤ 10. So probability 10 E[X] = 10 X iP (X = i) = i=1 10 X i· i=1 11 1 = 10 2 (b) Suppose we have the following strategy: if there are n remaining numbers a1 , . . . , an we ask ”Is it in the set {a1 , . . . , abn/2c }?”. For example, to find 1 first ask ”Is it in the set {1, 2, 3, 4, 5}?” and get answer ”Yes”. Then ask ”Is it in the set {1, 2}?” and get answer ”Yes”. Finally ask ”Is it in the set {1}?” and get answer ”Yes”. So in this case X = 3. Similarly to find 9 we can ask ”Is it in the set {1, 2, 3, 4, 5}?” get ”No”; ask ”Is it in the set {6, 7}?” get ”No”; ask ”Is it in the set {8}?” get ”No”; ask ”Is it in the set {9}?” and get ”Yes”. So in this case X = 4. With this strategy one needs 3 questions to find 1, 2, 3, 6, 7, 8 and 4 questions to find 4, 5, 9, 10. Thus E[X] = 3P (X = 3) + 4P (X = 4) = 3 · 4 17 6 +4· = 10 10 5 Problem 4.33 Suppose newsboy purchases m paper. Since the demand is binomial with n = 10 obviously profit will be maximized for some m ≤ 10. Let P be his profit and D be the demand in a day. Note that he pays total 10m cents for m purchased newspapers and sells min{D, m} of them. So E[P ] = 10 X 15 min{k, m}P (D = k) − 10m = 15 k=0 m−1 X kP (D = k) + 15m k=0 10 X P (D = k) − 10m k=m Let f (m) be right hand side of above equation. Then f (m + 1) − f (m) = 15 m X kP (D = k) − 15 kP (D = k) k=0 k=0 + 15(m + 1) 10 X P (D = k) − 15m k=m+1 10 X = 15 m−1 X 10 X P (D = k) − 10 = 15 1 − k=m+1 m X = 5 − 15 P (D = k) − 10 k=m m X ! P (D = k) k=0 P (D = k) . k=0 So we just need to find the maximum value of m such that m m k 10−k X X 10 1 2 1 FD (m) = P (D = k) = ≤ k 3 3 3 k=0 k=0 2 − 10 Observe that FD (2) = 1 23 × 28 < 39 3 FD (3) = 1 43 × 28 > 9 3 3 but 1 1 for all m ≤ 2 and FD (m) > 3 3 for all m ≥ 3. Hence f (m + 1) − f (m) > 0 for all m = 0, 1, 2 and f (m + 1) − f (m) < 0 for all m = 3, . . . , 9. This means f gets its maximum when m = 3 i.e., newsboy needs to buy 3 newspapers to maximize his profit. Since FD is an increasing function we conclude that FD (m) < Problem 4.42 (a) Let Xi = 1 if A answers i’th question correctly and Xi = 0 otherwise. Let Yi = 1 if B 10 X Xi Yi is the total number answers i’th question correctly and Yi = 0 otherwise. Then Z = i=1 of questions answered by both of A and B. Hence E[Z] = 10 X E[Xi Yi ] = i=1 10 X E[Xi ]E[Yi ] = i=1 10 X 0.7 · 0.4 = 2.8 i=1 (b) Now let Xi0 = 1 if A’s answer for i’th question is wrong and Xi0 = 0 otherwise. Similarly, let 10 X (1−Xi Yi ) = Yi = 1 if B’s answer for i’th question is wrong and Yi = 0 otherwise. Then Z = i=1 10 − 10 X Xi Yi is the total number of questions answered by at least one of A and B. Hence i=1 Var(Z) = Var 10 X i=1 ! Xi Yi = 10 X Var(Xi Yi ) = i=1 10 X 03 · 0.6 − (03 · 0.6)2 = 1.476 i=1 Problem 4.78 For 1 ≤ i ≤ n let Xi be the eventthat selection does not have exactly 2 the outcome of the i’th 4 4 T 17 n−1 c white balls. Then P (Xi ) = 1 − 2 82 = and A = i=1 Xi ∩ Xn is the event that we shall 35 4 make exactly n selections. Then P (A) = P (Xnc ) n−1 Y P (Xi ) = i=1 3 18 35 17 35 n−1 Problem 4.T.13 We want to maximize f (p) = P (X = k) = n k p (1 − p)n−k for 0 ≤ p ≤ 1. Since k n kpk−1 (1 − p)n−k − pk (n − k)(1 − p)n−k−1 f (p) = k n k−1 = p (1 − p)n−k−1 (k(1 − p) − (n − k)p) k n k−1 = p (1 − p)n−k−1 (k − np) k 0 f is increasing for p < k k k and decreasing for p > . So f (p) maximizes when p = . n n n Note: Another interesting question would be to find k which maximizes P (X = k) for a fixed n and k n k p n−k n n p. To find such k write P (X = k) = p (1 − p) = (1 − p) . So it is enough 1−p k k k p n . Observe that to maximize g(k) = 1−p k k+1 k n p p p n−k n−k n p = · = g(k) · g(k + 1) = k 1−p 1−p 1−p k+1 1−p k+1 k+1 So p n−k · ≥ 1 ⇐⇒ (n + 1)p − 1 ≥ k 1−p k+1 So g(k) achieves its maximum when k = b(n + 1)pc. g(k + 1) ≥ g(k) ⇐⇒ Problem 4.T.16 Observe that λi λi−1 λ λ = e−λ · = P (X = i − 1) · i! (i − 1)! i i So P (X = i) is increasing for i ≤ λ and decreasing for i ≥ λ. So it gets maximum when i = bλc. P (X = i) = e−λ Problem 4.T.21 (a) Let Ai be the event that persons 3 and 4 have birthdays on day i and let Bj be the event that persons 1 and 2 have birthdays on day j. Then 365 X 365 X P (E3,4 |E1,2 ) = P (E3,4 E1,2 ) = P (E1,2 ) i=1 j=1 365 X j=1 4 365 X 365 X P (Ai Bj ) P (Bj ) = i=1 j=1 365 X j=1 1 3654 1 3652 = 1 365 (b) Let Ck be the event that persons 1, 2 and 3 have birthdays on day k. Then 365 X P (E1,3 |E1,2 ) = P (E1,3 E1,2 ) = k=1 365 P (E1,2 ) X j=1 365 X P (Ck ) = P (Bj ) k=1 365 X j=1 1 3653 = 1 3652 1 365 (b) 365 X P (E2,3 E1,2 E1,3 ) = k=1 P (E2,3 |E1,2 E1,3 ) = 365 P (E1,2 E1,3 ) X P (Ck ) =1 P (Ck ) k=1 This shows that events Eij are pairwise independent but not mutually independent. 5 Problem 4.T.27 Since P (X > n) = ∞ X P (X = n + k) = k=1 ∞ X p(1 − p)n+k−1 = p(1 − p)n k=1 ∞ X (1 − p)k−1 = (1 − p)n k=1 we get P (X = n+k|X > n) = P (X = n + k) p(1 − p)n+k−1 P ({X = n + k}{X > n}) = p(1−p)k−1 = P (X = k) = = P (X > n) P (X > n) (1 − p)n This is because failing in the first n trials does not give any information about the future. In particular, after n fails the process of waiting for the first success will be as if you just started the game and waiting for the first success. 6