1 Mid-term Examination ST305 Statistical Inference 1. [2 marks] An experiment consists of ipping two coins. Write down the sample space for the outcome, which represents the results of the both coins. 1. Solution. The sample space S = {(h, h), (h, t), (t, h), (t, t)}, where t = tail and h = head. 2. [6 marks] Suppose the sample space is S = {(a, a), (b, b), (a, b), (b, a)}, and dene the events, A = {(a, b), (b, a)}, B = {(a, a)} and C = {(a, a), (b, b), (a, b)}. Find A ∩ B , A ∪ B , and C c . 2. Solution. A∩B =∅ A ∪ B = {(a, b), (b, a), (a, a)} C c = {(b, a)}. 3. [7 marks] Consider two independent events A and B . Suppose P (A) = 0.1 and P (B) = 0.2. Calculate P (Ac B c ). 3. Solution. A and B is independent =⇒ Ac and B c are also independent P (Ac B c ) = P (Ac )P (B c ) = (1 − P (A))(1 − P (B)) = (1 − 0.1)(1 − 0.2) = 0.72. 2 4. [6 marks] Suppose P {X = i} = n i 0.4i 0.610−i , i = 0, 1, . . . , n. Calculate P (X ≤ 1) and P (X ≥ 1). 4. Solution. P (X ≤ 1) = P (X = 0) + P (X = 1) = 0.40 0.610 + 100.40.69 = P (X ≥ 1) = 1 − P (X = 0) = 1 − 0.40 0.610 = 5. [10 marks] Suppose the random variable X has the following probability density function 1 f (x) = e−x + e−2x , x ≥ 0. 2 Calculate P (X > 5) and E[X]. 5. Solution. ∞ Z 5 1 1 1 ( e−x + e−2x )dx = e−5 + e−2∗5 = 2 2 2 Z ∞ P (X > 5) = 1 x( e−x + e−2x )dx 2 Z0 ∞ Z ∞ 1 −x = x e dx + e−2x )dx 2 0 Z0 ∞ Z ∞ 1 −x 1 = (−x) de + (− )de−2x 2 2 0 0 1 1 = + 2 4 E[X] = 6. [5 marks] Let X be a Poisson random variable with mean 5 and variance 10. Suppose Y = 2X + 10. Calculate E[Y ] and V ar(Y ). 3 6. Solution. E[Y ] = E[2X + 10] = 2E[X] + 10 = 2 ∗ 5 + 10 = 20 V ar(Y ) = V ar(2X + 10) = 4V ar(X) = 4 ∗ 10 = 40 7. [6 marks] Let Z be a standard normal random variable. Calculate (a) P (X ≤ 2), and (b) P (X ≤ −2). 7. Solution. (a) (b) P (X ≤ −2) = 1 − P (X ≤ 2) = 8. [12 marks] Suppose X is a random variable with the following probability density function: f (x) = (x−5)2 1 √ e− 2×100 , 10 2π −∞ < x < ∞ Calculate P (5 ≤ X ≤ 15). 8. . Solution. X ∼ N (5, 100) (X − 5)/10 ∼ N (0, 1) P (−1 ≤ X ≤ 1) = P (5 − 5)/10 ≤ X ≤ (15 − 5)/10) = P (0 ≤ X ≤ 1) = P (X ≤ 1) − P (X ≤ 0) = 9. (a) [10 marks] Let X be a random variable with a probability mass function: −2 n fX (n) = e 2 , n! n = 0, 1, . . . . Show that the moment generating function of X is MX (t) = e2(e −1) . Provide the full details of derivation. t 4 (b) [9 marks] Suppose the moment generating function of a random variable t Y is MY (t) = e2(e −1)+2t . Find the probability density/mass function of Y . 9. Solution. MX (t) = E[etX ] ∞ X = etn fX (n) = 0 ∞ X 0 ∞ X −2 n tn e 2 e n! t t e−2+2e e−2e (2et )n = n! 0 P t ∞ −2et (2et )n e−2+2e 0 e = n! −2+2et =e . (b) MY (t) = e2(e −1) e2t = MX (t)e2t =⇒ Y = X + 2. Thus, t fY (n) = P (Y = n) = P (X + 2 = n) = P (X = n − 2) e−2 2x , n = 2, 3, . . . = (n − 2)! 10. [8 marks] Joe is 80% certain that his missing remote controlled is in one of the two drawers of his TV cabinet, being 30% certain it is in the left drawer and 50% certain it is in the right drawer. If a search of the left drawer does not nd the remote controller, what is the conditional probability that it is in the right drawer? 5 10. Solution. L: the event that the it is in the left drawer R: the event that it is in the right drawer The desired probability: P (R|Lc ) = P (R) 50% 5 P (RLc ) = = = . P (Lc ) 1 − P (L) 1 − 30% 7 11. Suppose the moment generating function of a random variable X is MX (t) = 0.1 1 − 0.1t 4 if t < 10. Calculate E[X] and V ar(X). 11.Solution. E[X] = E[X 2 ] = d MX (t) dt d2 MX (t) dt2 = t=0 = t=0 4 ∗ 0.1 (1 − 0.1t)5 = 0.4 t=0 4 ∗ 5 ∗ 0.1 ∗ 0.1 (1 − 0.1t)6 = 0.2. t=0 V ar(X) = E[X 2 ] − (E[X])2 = 0.2 − 0.42 = 0.04. 11. Solution. 2 E[X] = m′ (t)|t=0 = 2tet |t=0 = 0 2 E[X 2 ] = m′′ (t)|t=0 = 4t2 e−t |t=0 = 0 12. [9 marks] Recall that family of pdfs/pmfs is called an family if it has the form: f (x; θ) = h(x)c(θ) exp k X i=1 ! wi (θ)ti (x) IA (x). exponential 6 Consider the distribution with its probability density function given by: g(x; θ) = θ−1 exp(1 − (x/θ)), 0 < θ < x < ∞. Show whether this distribution belong to the exponential family for not. Please provide the full details of derivation. 12. We can write Solution. x f (x; θ) = θ−1 exp 1 − I[θ,∞) (x) θ = h(x)c(θ) exp(w(θ)t(x))I[θ,∞) (x), where h(x) = e1 , c(θ) = θ−1 , w(θ) = θ−1 , The indicator function I[θ,∞) (x): • is a function of both θ and x =⇒ Thus, this is NOT an exponential family. t(x) = −x.