ISyE 6412A Theoretical Statistics Fall 2019
Instructor: Dr. Yajun Mei, Groseclose 343. Email: ([email protected]). Phone: 404-894-2334
Class Meets: TR 8:00-9:15am in IC 209. See Canvas for class homepage
Office Hours: After the classes, i.e., TR 9:15-9:45am in Groseclose 343.
Textbook: Recommended: “Statistical Inference,” by George Casella and Roger L. Berger; Duxbury
Press, Second Edition. ISBN: 0-534-24312-6. Additional notes on Statistical Decision The-
ory will be provided. All lecture notes will also be posted at Canvas.
Pre-requisites: ISYE 2027 & 2028 (or equivalent); Solid math skills (especially calculus). The students’
probability backgrounds are expected to be at the level of Ch. 1-5 of the textbook.
Homeworks: Weekly Homeworks will be assigned, and due back at Canvas after one week (we will
use Canvas extensively to save trees). Active, live collaboration/discussion is allowed and
encouraged on the homework, but each student must write down the solution in her
or his own individual way and there should be no two identical solutions to any problem.
In particular, do not consult the solution sets from classmates, previous years or
online when working this year’s problems. Late homework might be accepted with 50%
penalty within 72 hours of the deadline (and no penalty if you have a valid reason, i.e., a
note from the Dean’s Office). No homework is accepted after 72 hours of the deadline.
Exams: All exams are closed book. Students are allowed to bring two (2) cheat sheets during
Midterm I, four (4) cheat sheets during Midterm II, and six (6) cheat sheets during
the final exam (both sides are fine). The two midterm exams will be seventy-five (75)
minutes (during 8:00-9:15am), and the final exam will be longer (during 8:00-10:50am), as
the final exam covers all materials that have been taught in class. No make-up exam will
be given unless prior arrangement is made with the instructor with a valid reason, e.g., a
note from the dean.
Midterm I: Sep. 26 (Thursday)
Midterm II: Nov. 07 (Thursday)
Final: Dec. 12 (Thursday), 8:00-10:50am
Grading: The course grade is based on homeworks (20%), Midterms (20% + 20%), and Final (40%).
Software: If necessary, R statistical software package will be used for the examples discussed in class.
Handouts with R code lines will be provided from time to time.
TA: Ms. Min Fen, ([email protected]). Office hours: 12:30-1:30pm on Mondays and 1:30-
2:30pm on Wednesdays, ISyE Main #343.
Honor Code: It is your responsibility to get familiar with the Georgia Tech Honor Code. In particu-
lar, it is cheating if you copy homework/exam solutions from your classmates, previous
years, or online, and the penalty will include a lower course grade or even Fail. For any
questions involving these or any other Academic Honor Code issues, please consult me or
http://osi.gatech.edu/content/honor-code
Accommodation: At Georgia Tech we strive to make learning experiences as accessible as possible, and your
experience in this class is important to me. If you have already established accommodations
with the Offices of Disability Services, please communicate your approved accommodations
to me at your earliest convenience so we can discuss your needs in this course.
If you have not yet established services through Disability Services, but have a tempo-
rary health condition or permanent disability that requires accommodations (conditions
include but not limited to; mental health, attention-related, learning, vision, hearing, phys-
ical or health impacts), you are welcome to contact Disability Services at 404.894.2563 or
[email protected]. See more information at disabilityservices.gatech.edu.
Please turn page over →
Table 1: Tentative Schedule (the actual schedule may vary)
week Date Lect# Content
1 Aug 20 1 Introduction; Specification & Classification of a Statistical Problem
22 2 Statistical Decision Theory (SDT): Risk function; Admissible
2 Aug 27 3 The Bayes Paradigms
29 4 Focusing on Point Estimation
3 Sep 03 5
05 6 The Minimax Paradigms
4 Sep 10 7
12 8 Relationship between Admissiblity, Bayes and Minimax
5 Sep 17 9
19 10 SDT for high-dim data: James-Stein estimator (Shrinkage)
6 Sep 24 11
26 12 Midterm I
7 Oct 01 13 Sufficient Statistics: Sufficient; Factorization Theorem
03 14 Minimal Sufficient; Ancillary
8 Oct 08 15 Complete Sufficient; Basu’ Theorem
10 16
9 Oct 15 no class, fall break
17 17
10 Oct 22 18 Point Estimation (Ch. 7, 10): MOM; MLE
24 19 UMVUE (Uniformly Minimum Variance Unbiased)
11 Oct 29 20 Rao-Blackwell Method; Cramer-Rao Inequality
31 21
12 Nov 05 22
07 23 Midterm II
13 Nov 12 24 Asymptotic Properties of MLE (Ch. 10)
14 25
14 Nov 19 26 Modern applications: sparsity in high-dim data — LASSO
21 27
15 Nov 26 28 Hypothesis Testing & Classification in Machine Learning (Ch. 8, 9, 10)
28 no class, thanksgiving
16 Dec 03 29 Last Day of Classes
Final Dec 12 Final Exam: (Thursday, Dec 12) 8:00-10:50am
