STA347H1F Term Test 1 (LEC 5201) October 19th, 2021 6:10pm - 7:40pm EST All relevant work must be shown for full marks Total Marks = 40 Need to show all your work to receive full marks Note: There are 4 questions (with parts) and you have to answer all of them to receive full marks. The answers can be hand written or typed. You can submit multiple files but please make sure all the answers are submitted in pdf, docx, jpg or png files. You will not be able to submit anything after 7:40pm EST (other than a few exceptions). The question are created in a way that all of them can be answered within 1 hours and 15 minutes. Please make sure that you have enough time to scan and submit. ACADEMIC INTEGRITY: This is an open book exam. However, the University treats cases of plagiarism and cheating very seriously. It is the students’ responsibility for knowing the content of the University of Toronto’s Code of Behaviour on Academic Matters. All sus- pected cases of academic dishonesty will be investigated following procedures outlined in the above document. If you have questions or concerns about what constitutes appropriate academic be- haviour or appropriate research and citation methods, you are expected to seek out additional information on academic integrity from your instructor or from other institutional resources (see http://academicintegrity.utoronto.ca/). Here are a few guidelines regarding academic in- tegrity: • You may consult class notes/lecture slides during the test, however sharing or discussing questions or answers with other students is an academic offense. • Students must complete all assessments individually and independently. Working together is not allowed. • Paying anyone else to complete your assessments for you is serious academic misconduct. • Sharing your answers/work/code with others is academic misconduct. • Looking up solutions to test/quiz problems online and copying what you find is an academic offense. • All work that you submit must be your own! You must not copy mathematical derivations, computer output and input, or written answers from anyone or anywhere else. Unacknowl- edged copying or unauthorized collaboration will lead to severe disciplinary action, beginning with an automatic grade of zero for all involved and escalating from there. Please read the UofT Policy on Cheating and Plagiarism, and don’t plagiarize. Please don’t upload this document in any social media platforms, chegg, slideshare or coursehero. Uploading this document to any such website will be treated as a serious academic offense and we will take actions based on University of Toronto’s policies regarding plagiarism. We will constantly keep an eye in these websites to find out such incidences. Best of luck! 1 Question 1 (a) [2 Marks] Let X ∼ Geometric(θ), x = 0, 1, 2, ..... Calculate the MGF of X Solution: The MGF, mX(t) = E(e tX) = ∞∑ x=0 etx(1− θ)xθ = θ ∞∑ x=0 ( et(1− θ))x = θ 1− et(1− θ) where, et(1− θ) < 1 or t < − log(1− θ) (b) [4 Marks] Let Y ∼ Negative Binomial(r, θ). Calculate the MGF of Y . (P (Y = k) = (r+k−1r−1 )θr(1− θ)k; k = 0, 1, 2, ...) Solution: There are two ways to do that. The easier option is to assume Y = X1+X2+ ...Xr, where X1, X2, ..., Xr are independent and have geometric (θ) distributions. Thus, Y = X1 + X2 + ...Xr ∼ NB(r, θ). Thus, the MGF, mY (t) = (mX(t))r = ( θ 1− et(1− θ) )r . The second way is to use the negative-binomial formula. That is, mY (t) = E(e tY ) = ∞∑ y=0 ety ( r + y − 1 r ) θr(1− θ)y = θr ∞∑ y=0 ( r + y − 1 r )( et(1− θ))y = θr (1− et(1− θ))r again, for et(1− θ) < 1 or t < − log(1− θ). (c) [6 Marks] Using the MGF of negative binomial distribution from (b) calculate the mean and the variance of Y . Solution: The first derivative of the MGF, m′Y (t) = −r(−et)(1− θ)θr (1− et(1− θ))r+1 m′Y (0) = r(1− θ)θr (θ)r+1 E(Y ) = r (1− θ) θ The second derivative of the MGF, m′′Y (t) = ret(1− θ)θr (1− et(1− θ))r+1 + r(r + 1)e2t(1− θ)2θr (1− et(1− θ))r+2 ⇒ m′′Y (0) = E(Y 2) = r(1− θ)θr θr+1 + r(r + 1)(1− θ)2θr θr+2 = r(1− θ)(1 + r − rθ) θ2 2 Thus, the variance, V (Y ) = E(Y 2)− (E(Y ))2 = r(1− θ)(1 + r − rθ) θ2 − ( r (1− θ) θ )2 = r(1− θ) + r2(1− θ)2 θ2 − r 2(1− θ)2 θ2 = r(1− θ) θ2 Question 2 x y A O r B Z θ The above circle is a unit circle. That is the radius of the circle is r = 1 unit. The triangle OAB is right triangle with the angle ∠OBA = pi2 . The angle θ is a random variable which follows Uniform(0, pi2 ). Let, Z be the length of AB, which depends on θ. (a) [5 Marks] Calculate E(eθ). Find its relation with eE(θ) (i.e., equal, smaller or larger). Solution: Since θ has a Uniform(0, pi2 ) distribution, hence, E(θ) = pi4 and eE(θ) = e pi 4 . Now, E(eθ) = ∫ pi 2 0 eθ 1 pi 2 dθ = 2 pi [ eθ ]pi 2 0 = 2 pi ( e pi 2 − 1 ) Since, eθ is a convex function, from Jensen’s inequality we can see that eE(θ) < E(eθ). Since, we know that it is not a linear function. Note: Or one can just calculate the values that E(eθ) = 2.425825 and eE(θ) = 2.19328. Hence, eE(θ) < E(eθ) (b) [5 Marks] Show that E(Z) < 1√ 2 . Solution: We need to identify that Z = sin(θ). Since sin(θ) is a concave function between 0 to pi/2, thus, E(Z) = E(sin(θ)) ≤ sin(E(θ)) = sin(pi/4) = 1√ 2 , by using Jensen’s inequality. 3 Question 3 A man is given n different keys, only one of which will open his door. He tries them successively until he can open the door and then he stops. (a) [3 Marks] Describe the sample space of this experiment. Solution: There are n keys in total. The possibilities are S, FS, FFS, FFFS, ...,, where the sequence ends on the nth try. Let’s define Si = { 1, if the door opens at the ith try 0, if the door does not open at the ith try Thus, the sample space = {S1, S2, ..., Sk|k ≤ n, Sk = 1} (b) [5 Marks] Show that each outcome of the experiment has the same probability. What is the probability? Solution: The probability that he will be successful at the first try, P (S) = 1n . The probability that he will be successful at the second try is P (FS) = n−1n 1n−1 = 1n . The probability of success on the third try P (FFS) = n−1n n−2n−1 1n−2 = 1n . As we can see that ∀k ≤ n, P (Sk) = 1n . Question 4 (a) [3 Marks] Consider, for α > 0 fixed, a function is given by f(x) = α(1+x)−α−1, for 0 < x <∞. Show that f(.) is a density. Solution: We have that f(x) ≥ 0 for every x. We have,∫ ∞ 0 α(1 + x)−α−1dx = [ − (1 + x)−α ]∞ 0 = 1 Thus, f(x) is a density. (b) [3 Marks] Calculate the CDF of the distribution provided in (a). Solution: The CDF can be given as, F (x) = ∫ x 0 α(1 + u)−(α+1)dz = 1− (1 + x)−α (1) (c) [3 Marks] Calculate the median of the distribution given in (a). Solution: From the CDF we can calculate, 1− (1 +M)−α = 1 2 ⇒(1 +M)−α = 1 2 ⇒1 +M = 2 1α ⇒M = 2 1α − 1 4
欢迎咨询51作业君