Scribd Logo
Upload

Welcome to Scribd!

  • 1 views

    STAT 709 Midterm 2024

    Uploaded by

    soumengosh404

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    STAT 709 Midterm 2024

    Uploaded by

    soumengosh404
    1 views2 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    1 views2 pages

    STAT 709 Midterm 2024

    Uploaded by

    soumengosh404

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 2

    STAT 709: Midterm Exam

    Instructor: Yiqiao Zhong, TA: Jingyang Lyu


    Fall 2024

    Date: Oct. 28, 2024, 11:00 am—12:15 pm

    1. (40 points, 8 points for each subproblem) Suppose that X ∼ N (0, 1) is a standard normal random variable.
    Denote the cumulative distribution function (CDF) of X by Φ(x). Let Y be a “censored” random variable
    of X, defined by 
    a,
     if X ≥ a
    Y = X, if |X| < a, (1)

    −a, if X ≤ −a

    where a > 0 is a given scalar. Let δx denote the delta measure defined as
    (
    0, if x ∈
    /A
    δx (A) = for any Borel set A
    1, if x ∈ A

    (i) Find the CDF of Y using Φ.


    (ii) Find the CDF associated with the measure δa .
    (iii) Let µX and µY be induced measure of X and Y respectively. Express µY in terms of µX and delta
    measures.
    (iv) Does µY have (Lebesgue) probability density function, or probability mass function? Briefly explain
    why.
    (v) Let sign(x) be the sign function defined by

    1,
     if x > 0
    sign(x) = 0, if x = 0

    −1, if x < 0

    Show that |Y | and sign(Y ) are independent.

    2. (40 points, 8 points for each subproblem) Suppose that we have i.i.d. k-dimensional observations
    X1 , X2 , . . . , Xn ∼ N (0, Σ) where Σ = diag{σ12 , . . . , σk2 } is an unknown diagonal matrix with σ12 , . . . , σk2 > 0.
    We are interested in estimating Σ using the n observations.
    (i) Do we have a parametric distribution or nonparametric distribution?
    (ii) Write down the likelihood function for estimating Σ.
    (iii) Find the MLE Σ b for Σ.
    (iv) What is the bias of the estimator Σ? b
    (v) Suppose that we are now interested in estimating Σ1/2 = diag{σ1 , . . . , σk }. Denote Σ b 1/2 = diag{b bk }.
    σ1 , . . . , σ
    Here the square root means taking the square root of each element in the diagonal matrix. Use the Delta
    method to derive √  d
    n (b bk ) − (σ1 , . . . , σk ) −→ N (0, V )
    σ1 , . . . , σ

    1
    and determine an expression for the matrix V ∈ Rk×k .

    3. (20 points, 10 points for each subproblem) Let G ∼ N (0, 1) be a standard normal variable, and
    f : R → R be a continuously differentiable function.
    (i) Prove Stein’s identity: E[f 0 (G) − Gf (G)] = 0 where f and f 0 are bounded.
    (ii) Let (Xn )n≥1 be a sequence of random variables with supn E[Xn2 ] ≤ C < ∞. Stein’s continuity theorem
    is a useful result, which states that the following two statements are equivalent.
    (a) E[f 0 (Xn ) − Xn f (Xn )] converges to zero whenever f : R → R is continuously differentiable with f, f 0
    bounded.
    (b) Xn converges in distribution to G.

    Please prove that (b) implies (a).

    You might also like