Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
STATS 210
1. Suppose that the random variables X and Y have the following joint pdf: fX,Y (x, y) = 2 for 0 < x < 1 and 0 < y < 1− x 0 otherwise Answer the following questions: (a) Find the marginal pdf of Y . For writing down your result, replace . . . with the correct expression: fY (y) = { . . . for 0 < y < 1 0 otherwise [3 marks] (b) Show that the pdf of X given Y has the following expression for 0 < y < 1: fX|Y (x|y) = 1 1− y for 0 < x < 1− y 0 otherwise [1 mark] (c) Use the result obtained in part (b) to show that E(X|Y = y) = 1− y 2 for 0 < y < 1 [3 marks] (d) Employ the result obtained in part (c) to write down the expression of E(X|Y ) and then use this expression for calculating the ratio Cov(X,Y ) Var(Y ) without computing Cov(X,Y ) and Var(Y ). Assume that 0 < Var(Y ) <∞. [3 marks] (e) Decide if the random variables X and Y are independent (or not). Justify your answer. [1 mark] [Total: 11 marks] 1 STATS 210 2. Suppose that each of two statisticians A and B must estimate a certain parameter θ whose value is unknown (θ > 0). Statistician A can observe the value of a random variable X, which has the Poisson distribution with mean 2θ. The value observed by statistician A is X = 3. (a) Write down the likelihood function for this problem, LA(θ; 3). Remember to state the range of values of θ for which the likelihood is defined. Find the maximum likelihood estimate of θ. In your answer, you should differentiate the likelihood with respect to the parameter θ, and set to 0 for the maximum. Quote your numerical answer to 1 decimal place. You should make reference to the graph of the likelihood function, which is shown in Figure 1. Show all your working in your answer. [5 marks] 0 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25 L A (; 3) Figure 1: Plot for question 2, part (a). Question 2 continues on the next page 2 STATS 210 Statistician B can observe the value of a random variable Y ∼ Gamma(3, θ). Suppose that the value observed by statistician B is Y = 2. (b) Write down the log-likelihood function for this problem, logLB(θ; 2). Remember to state the range of values of θ for which the log-likelihood is defined. Find the maximum likelihood estimate of θ. In your answer, you should differentiate the log-likelihood with respect to the parameter θ, and set to 0 for the maximum. Quote your numerical answer to 1 decimal place. You should make reference to the graph of the log-likelihood function, which is shown in Figure 2. Show all your working in your answer. [5 marks] 0 1 2 3 4 5 -14 -12 -10 -8 -6 -4 -2 0 lo gL B( ;2 ) Figure 2: Plot for question 2, part (b). (c) Use the results obtained in part (a) and in part (b) to show that there exists a constant C (which does not depend on θ) such that logLB(θ; 2) = logLA(θ; 3) + C for all θ > 0. Find the numerical value of C. Quote your answer to 4 decimal places. [2 marks] (d) Compare the values of the maximum likelihood estimates that you have obtained in part (a) and in part (b). Use the identity from part (c) in order to justify why the maximum likelihood estimate obtained by statistician B should be the same as the maximum likelihood estimate obtained by statistician A. Explain your answer. [5 marks] [Total: 17 marks] 3 STATS 210 3. Suppose that either Instrument A or Instrument B might be used for making a certain measurement. From inspecting visually the instrument, we cannot distinguish between Instrument A and Instrument B. Instrument A yields a measurement whose (marginal) pdf is: fX(x) = 2x for 0 < x < 1 0 otherwise Instrument B yields a measurement whose (marginal) pdf is: fY (y) = 3y2 for 0 < y < 1 0 otherwise (a) Find the (marginal) cdf of X. For writing down your result, replace . . . with the correct expressions: FX(x) = · · · for x ≤ 0 · · · for 0 < x < 1 · · · for x ≥ 1 [4 marks] (b) Find the (marginal) cdf of Y . For writing down your result, replace . . . with the correct expressions: FY (y) = . . . for y ≤ 0 . . . for 0 < y < 1 . . . for y ≥ 1 [4 marks] Suppose that one of the two instruments is chosen uniformly at random and a measure- ment T is made with it. Note that T is a continuous random variable. Let I be a Bernoulli random variable with the property that I = 0 if Instrument A is chosen and I = 1 if Instrument B is chosen. It is evident that P(I = 0) = P(I = 1) = 1/2. (c) Assume that 0 < t < 1. Fill in the gaps indicated with . . . in the following identity: P(T ≤ t) = P(T ≤ t| . . .)P(. . .) + P(T ≤ t| . . .)P(. . .). [2 marks] Question 3 continues on the next page 4 STATS 210 (d) Use the results obtained in parts (a)-(c) in order to show that the expression of the (marginal) cdf for T is: FT (t) = 0 for t ≤ 0 t2 + t3 2 for 0 < t < 1 1 for t ≥ 1 [3 marks] (e) Find the (marginal) pdf of T . For writing down your result, replace . . . with the correct expressions: fT (t) = { . . . for 0 < t < 1 . . . otherwise [3 marks] (f) Assume that 0 < t < 1. Fill in the gaps indicated with . . . in the following identity: P(I = 1 |T = t) = f...|...(t | I = 1)P(. . .) f...(. . .) Name the theorem that is the basis for the identity above. [3 marks] (g) Show that P (I = 0 |T = 1/4) = 8/11. [3 marks] [Total: 22 marks] 5 STATS 210 4. Let U ∼ Uniform(0, 1). We define T = log U 1− U = log(U)− log(1− U). Answer the following questions: (a) Verify that the function g(u) = log(u) − log(1 − u) is (strictly) monotone increasing by showing that dg(u) du > 0 for 0 < u < 1. [1 mark] (b) Show that lim u→0 g(u) = −∞ and lim u→1 g(u) =∞. [1 mark] (c) Find the pdf of T , remembering to include the range of values of t for which your answer is valid. [3 marks] [Total: 5 marks] 5. Suppose that the distribution of (X,Y ) is Bivariate Normal, the marginal distribution of X is Normal(0, 1) and the marginal distribution of Y is Normal(0, 9). The correlation between X and Y is ρX,Y = −1 3 . We define: V = X, W = −kX + Y, where k is a constant. Answer the following questions: (a) Name the joint distribution of (V,W ). Do not use calculus. [1 mark] (b) Write down the expression of Cov(V,W ). [2 marks] (c) Find the value of k such that the random variables V and W are independent. [2 marks]