Quiz 5B for MATH 105 SECTION 205 March 20, 2015 Given Name Family Name Student Number 1. Let f (x) = ae−2x , if x ≥ 1, for some constant a such that f (x) is a PDF for some continuous random 0, otherwise variable X. (a) (0.5 points) Find a. 2e2 (a) Z Answer. ∞ f (x) dx = 1, that is, Since f (x) is a PDF, then −∞ ∞ Z f (x) dx 1 = Z−∞ ∞ = 1 Z ae−2x dx ∞ e−2x dx 1 1 −2x ∞ a· − e 2 1 a −2 e . 2 = a = = Then a = 2e2 . (b) (0.5 points) Find the cumulative distribution function F (x) of X. (b) Z Answer. F (x) = x Since F (x) = f (t) dt, then −∞ a. If x < 1, then Z x F (x) = Z x f (t) dt = −∞ 0 dt = 0. −∞ 0, if x ≤ 1, 1 − e2−2x , if x > 1. b. If x ≥ 1, then x Z f (t) dt F (x) = −∞ Z 1 Z = x f (t) dt + −∞ Z 1 x Z 0 dt + = −∞ 2 f (t) dt 1 2e2 e−2t dt 1 Z x = 2e e−2t dt 1 = 1 −2t x 2e · − e 2 1 2 = e2 (−e−2x + e−2 ) = −e2−2x + e2−2 = 1 − e2−2x . In summary, we get F (x) = 0, if x ≤ 1, 2−2x 1−e , if x > 1. (c) (0.5 points) Compute Pr(X ≤ 2). (c) Answer. 1 − e−2 In fact, we have Pr(X ≤ 2) = F (2) = 1 − e2−2·2 = 1 − e−2 . (d) (0.5 points) Compute the mean of X. (d) Answer. In fact, we have Z ∞ E(X) = xf (x) dx Z−∞ ∞ x · 2e2 e−2x dx 1 Z ∞ 2 = 2e xe−2x dx. = 1 Z For xe−2x dx, use the integration by parts, by ’ILATE’ rule, we have u = x =⇒ du = dx Z e−2x −2x dv = e dx =⇒ v = e−2x dx = − . 2 3 2 Then we have Z xe−2x dx = = = = = Z u dv Z uv − v du −2x Z −2x e e + dx x· − 2 2 Z xe−2x 1 − + e−2x dx 2 2 xe−2x e−2x − + C. − 2 4 Notice that x e2x 1 = lim x→∞ 2e2x = 0. lim xe−2x = x→∞ lim x→∞ By l’Hospital’s rule Hence, we get E(X) = = = = ∞ xe−2x e−2x 2e − − 2 4 1 e−2 e2 · e−2 + 2 −2 3e e2 · 2 3 . 2 2 (e) (0.5 points) Compute the variance of X. (e) Answer. Notice that Var(X) = E(X 2 ) − [E(X)]2 . Since Z ∞ x2 f (x) dx E(X 2 ) = Z−∞ ∞ = x2 · 2e2 e−2x dx 1 Z ∞ 2 = 2e x2 e−2x dx. 1 Z For x2 e−2x dx, use the integration by parts, by ’ILATE’ rule, we have u = x2 =⇒ du = 2xdx Z e−2x −2x dv = e dx =⇒ v = e−2x dx = − . 2 4 Then we have Z Z x2 e−2x dx = u dv Z = uv − v du −2x Z −2x e e 2 + · 2x dx = x · − 2 2 Z x2 e−2x = − + xe−2x dx 2 x2 e−2x xe−2x e−2x − − + C By the computations in part d = − 2 2 4 Notice that lim xe−2x = x→∞ = = lim x2 e−2x = x→∞ = = = x e2x 1 lim x→∞ 2e2x 0 x2 lim 2x x→∞ e 2x lim x→∞ 2e2x 2 lim x→∞ 4e2x 0. lim x→∞ By l’Hospital’s rule By l’Hospital’s rule By l’Hospital’s rule Hence, we get 2 E(X ) = = = = ∞ x2 e−2x xe−2x e−2x 2e − − − 2 2 4 1 −2 e e−2 e−2 2e2 · + + 2 2 4 −2 5e 2e2 · 4 5 . 2 2 Therefore, we get 2 2 5 3 25 9 16 − = − = = 4. Var(X) = E(X ) − [E(X)] = 2 2 4 4 4 2 2 (f) (0.5 points) Compute the standard deviation of X. (f) Answer. In fact, we have σ(X) = 2. For a sequence {an }∞ n=1 Z such that an = 0 1 √ p √ Var(X) = 4 = 2. dx . + x2 n2 2 (a) (0.5 points) Find the explicit formula of an . (a) Z Answer. For √ dx , let x = n tan(θ), then + x2 Z Z dx √ = n2 + x2 Z = n2 √ n2 + 1 1 ln + n n dx = n sec2 (θ) dθ and n2 + x2 = n2 sec2 (θ). So n sec2 (θ) dθ n sec(θ) sec(θ) dθ = ln | sec(θ) + tan(θ)| + C. Since tan(θ) = x , then n √ sec(θ) = So we get Z n2 + x2 . n √ n2 + x2 x dx √ = ln + + C. n n n2 + x2 Hence we have Z 1 an = √ 0 = = = = dx n2 + x2 √ 1 n2 + x2 x ln + n n 0 √ √ n2 n2 + 1 1 0 + − ln + ln n n n n √ n2 + 1 1 + − ln 1 ln n n √ n2 + 1 1 ln + . n n (b) (0.5 points) Find a100 . √ (b) Answer. ln 10001 + 1 100 By the result of part a, we have √ √ 1002 + 1 1 10001 + 1 a100 = ln + . = ln 100 100 100 (c) (0.5 points) If {an }∞ n=1 has a limit, find this limit. (c) Answer. In fact, since "√ lim n→∞ # n2 + 1 1 + = 1. n n 0 Then lim an = ln 1 = 0. n→∞ 3. For a sequence {an }∞ k=1 such that a1 = 1 and an+1 = 4an for all n ≥ 1. (a) (0.5 points) Compute a2 , a3 , a4 , a5 . (a) Answer. 4, 6, 64, 256 Since a1 = 1 and an+1 = 4an , then a2 = 4a1 = 4 · 1 = 4 a3 = 4a2 = 4 · 4 = 16 a4 = 4a3 = 4 · 16 = 64 a5 = 4a4 = 4 · 64 = 256 (b) (0.5 points) Find the explicit formula of an . (b) 4n−1 Answer. Since an+1 = 4an for all n ≥ 1, then {an } is a geometric sequence with ratio r = 4, which implies that an = a1 rn−1 = 1 · 4n−1 = 4n−1 . √ cos( n) 3 tan−1 (n) 4. (a) (0.5 points) Compute lim 2 · √ . + n→∞ n3 + 3 n (a) √ π π Notice that −1 ≤ cos( n) ≤ 1 and − ≤ tan−1 (n) ≤ , then 2 2 √ 1 tan−1 (n) π cos( n) π π −√ ≤ √ ≤ ≤ , ≤ √ , and − 3 3 3 2(n + 3) n +3 2(n + 3) n n n 0 Answer. for all n ≥ 1. π π Since lim √ = lim = 0, by the squeeze theorem, we have 3 n→∞ n n→∞ 2(n + 3) √ cos( n) tan−1 (n) √ lim = lim = 0. n→∞ n→∞ n3 + 3 n Hence we get √ cos( n) 3 tan−1 (n) √ lim 2 · + = 0. n→∞ n3 + 3 n n3 + 2n 23n−1 −1 (b) (0.5 points) Compute lim tan + n−10 . n→∞ n2 + 2n + 1 9 (b) π 2 n3 + 2n Answer. Since lim = ∞, then lim tan−1 n→∞ n2 + 2n + 1 n→∞ geometric sequence with ratio r: r = = = = Then lim n→∞ n3 + 2n n2 + 2n + 1 π = . For 2 23·2−1 91−10 · 92−10 23·1−1 25 1 · 22 9 23 9 8 < 1. 9 23n−1 = 0. 9n−10 Therefore, we get lim (c) (0.5 points) Evaluate k=5 tan n→∞ ∞ X −1 n3 + 2n n2 + 2n + 1 23n−1 π + n−10 = . 9 2 1 . (3k + 1)(3k + 4) (c) Answer. 1 48 By partial fraction decomposition, we have 1 1 1 1 1 = − = − . (3k + 1)(3k + 4) 3(3k + 1) 3(3k + 4) 3(3k + 1) 3[3(k + 1) + 1] So we have Sn = n+4 X k=5 1 (3k + 1)(3k + 4) 1 1 − 3(3 · 5 + 1) 3(3 · 6 + 1) 1 1 + − 3(3 · 6 + 1) 3(3 · 7 + 1) 1 1 + − 3(3 · 7 + 1) 3(3 · 8 + 1) .. . 1 1 − + 3(3 · (n + 3) + 1) 3(3 · (n + 4) + 1) 1 1 + − 3(3 · (n + 4) + 1) 3(3 · (n + 5) + 1) 1 1 = − 3(3 · 5 + 1) 3(3 · (n + 5) + 1) 1 1 1 = − → , as n → ∞. 48 3(3 · (n + 5) + 1) 48 = So we get ∞ X k=5 1 1 = lim Sn = . (3k + 1)(3k + 4) n→∞ 48 23n−1 , it’s a 9n−10 (d) (0.5 points) Evaluate ∞ X ln k=2 k+1 . k ∞ or divergent (d) Answer. Let Sn = n+1 X ln k=2 k+1 k+1 , notice that ln = ln(k + 1) − ln k, then k k Sn = ln 3 − ln 2 + ln 4 − ln 3 + ln 5 − ln 4 .. . + ln(n + 1) − ln(n) + ln(n + 2) − ln(n + 1) = ln(n + 2) − ln 2 → ∞, Then ∞ X k=2 ln as n → ∞. k+1 is divergent. k (e) (0.5 points) Evaluate ∞ X 3 · 42k−5 74k+5 k=8 . 3·411 737 3 1 − 474 (e) Answer. Notice that ∞ X 3 · 42k−5 k=8 74k+5 is a geometric series with initial term a: 3 · 42·8−5 74·8+5 3 · 411 . 737 a = = And the ratio r: r = = = 3 · 42·9−5 74·8+5 · 74·9+5 3 · 42·8−5 13 37 4 7 · 41 11 4 7 43 < 1. 74 So we get ∞ X 3 · 42k−5 k=8 (f) (0.5 points) Is the series ∞ X k=100 √ 74k+5 11 = 3·4 a 37 = 7 43 . 1−r 1 − 74 k+1 convergent or divergent? k (f) divergent √ Answer. Approach I: Let f (x) = x+1 = x r x+1 = x2 r 1 1 + , then f (x) is positive and decreasx x2 ing. Notice that ∞ Z Z ∞ f (x) dx = 1 Z √ For 1 √ x+1 dx. x x+1 dx, then x Z √ Z √ u x+1 dx = · 2u du let u = x + 1, then x = u2 − 1 and dx = 2udu 2 x u −1 Z u2 = 2 du u2 − 1 Z 1 du = 2 1+ 2 u −1 Z 1 = 2u + 2 du. 2 u −1 By the partial fraction decomposition, we have u2 1 1 1 = − . −1 2(u − 1) 2(u + 2) Then Z √ Z x+1 1 1 dx = 2u + 2 · − du x 2(u − 1) 2(u + 2) Z 1 1 = 2u + − du u−1 u+1 = 2u + ln |u − 1| − ln |u + 1| + C √ √ √ = 2 x + 1 + ln | x + 1 − 1| − ln | x + 1 + 1| + C √ x + 1 − 1 √ + C. = 2 x + 1 + ln √ x + 1 + 1 Notice that √ lim 2 x + 1 = ∞, x→∞ and √ x + 1 − 1 = ln 1 = 0. lim ln √ x→∞ x + 1 + 1 Then Z ∞ 1 √ x + 1 − 1 ∞ √ = ∞. f (x) dx = 2 x + 1 + ln √ x + 1 + 1 1 So by the integral test, we know that the series ∞ X k=100 Approach II: Notice that √ k+1 1 ≥ , k k √ k+1 is divergent. k for all k ≥ 1. ∞ X1 X Since diverges, by the comparison test, then the series k k=100 √ k+1 is divergent. k (g) (0.5 points) Is the series ∞ X 4 convergent or divergent? 2 k ln k k=9 convergent (g) Answer. Z For Let f (x) = 4 , then f (x) is positive and decreasing for x ≥ 9. Notice that x ln2 x Z ∞ Z ∞ 4 f (x) dx = dx x ln2 x 9 9 Z ∞ 1 = 4 dx. x ln2 x 9 1 dx, then x ln2 x Z Z du 1 Let u = ln x, then du = dx u2 x 1 = − +C u 1 + C Since u = ln x. = − ln x 1 dx = x ln2 x So we get Z ∞ f (x) dx = 9 = So by the integral test, we know that the series 1 ∞ 4 − ln x 9 4 . ln 9 ∞ X 4 is convergent. k ln2 k k=9 ax2 + b, if 0 ≤ x ≤ 1, for some constants a and b such that f (x) is a probability density 0, otherwise function for some continuous random variable X. 5. Let f (x) = (a) (1 point) Find conditions for a and b. Answer. Since f (x) is a PDF for some random variable, then f (x) ≥ 0 for all x. For 0 ≤ x ≤ 1, since f (x) = ax2 + b is a non-negative quadratic polynomial and symmetric with respect to x = 0, then Z ∞ f (1) = a + b ≥ 0 and f (0) = b ≥ 0. On the other hand, we must have f (x) dx = 1, then ∞ Z ∞ 1 = f (x) dx −∞ Z 1 = = = So we get (ax2 + b) dx 0 ha i1 x3 + bx 3 0 a + b. 3 a + b = 1, that is, a + 3b = 3. In summary, we must have the following conditions: 3 a + b ≥ 0, b ≥ 0, and a + 3b = 3. (b) (1 point) Compute a and b such that E(X) = 1. Answer. Since E(X) = 1, then 1 = E(X) Z ∞ xf (x) dx = −∞ 1 Z x(ax2 + b) dx = 0 Z = 1 (ax3 + bx) dx a 4 b 2 1 x + x 4 2 0 a b + . 4 2 0 = = So we get a b + = 1, that is, a + 2b = 2. By the result of part a, we have a + 3b = 3, then 4 2 a = 0, and b = 1. Your Score: /11