Instructor: Jimin Ding;
Office hours: Wed. 11:10 am -12:00pm. Fri. 4:10 pm.-5:30pm. or by appointment.
Topics covered:
Review of basic linear models relevant for the course; generalized linear models including logistic and Poisson regression (heterogeneous variance structure, quasilikelihood); linear mixed-effects models (estimation of variance components, maximum likelihood estimation, restricted maximum likelihood, generalized estimating equations), generalized linear mixed-effects models for discrete data, models for longitudinal data, optional multivariate models as time permits.
The computer software R will be used for examples and homework problems. Implementation in SAS will be mentioned for several specialized models.
Prerequisites:
Math 5071 and a course on mathematical statistics at the level of Math 494 (can be taken concurrently), or permission of instructor.
Reference Books:
- C.E. McCulloch, S.R. Searle, J. M. Neuhaus,
Generalized, Linear and Mixed Models, , 2nd ed.
John Wiley & Sons, 2008.
- W.W. Stroup, Generalized Linear Mixed Models: Modern Concepts, Methods and Applications,
Chapman & Hall/CRC, 2012.
- J. Jiang, Springer, Linear and Generalized Linear Mixed Models and Their Applications, 2010.
- G. Verbeke and G. Molenberghs, Springer, Linear Mixed Models for Longitudinal Data, 2000.
Exams and Homeworks:
There will be one in-class midterm on March 4, Wed., and the final exam on May 4, Mon., 3:30-5:30 pm. The take-home final project is due on May 2, Sat., 11:59pm. ( If you turn in your final within 24 hours after due date, the grade will be scaled by 70%. No final after May 3, 11:59pm. will be graded.
There will be one in-class presentation and homework sets every other week. Homework will be collected on Wednesday in class. Latehomework submitted within 2 days of due date will receive 25% penalty for each day late. Any homework late by more than 2 days will not be graded and receive zero point.
Remark:
Students who take this class as the statistics (or mathematics) Ph.D. qualifying exam subject will receive three hourfinal exam, and the decision on your qualifying exam is solely based on your performance on the final exam.
Grades:
Your grade will be based on the in-class presentation, the in-class midterm and the final exam, together with homeworks in the proportions. Then your final letter grade is determined as follows. The A range will be 85 to 100, the B range will be 70 to 85, the C range will be 60 to 70, and the D range will be 50 to 60, with plus and minus grades given to the top 10% and bottom 10% students in each of these ranges. (If you elect ``Credit/No Credit'', Cr means D or better.)
Homework and Projects |
40% |
Presentation |
15% |
Midterm exams |
20% |
Final exam |
25% |
Collaboration:
Collaboration on homework is allowed and can be helpful (and fun). However, you must do all written work by yourself, both answers to homework questions and computer programs. If you collaborate with someone on a homework, list his or her
name in a note at the top of the first part of your homework.
There should be NO COLLABORATION on exams.
Good books for reviewing elementary statistics:
- A Data-Based Approach to Statistics,R. L. Iman,
Duxbury Press, 1994.
- Statistics and Data Analysis
from Elementary to Intermediate, A. J. Tamhane and D. D. Dunlop, Prentice-Hall, 2000.
- Design and analysis of experiments, 2nd ed., Douglas
Montgomery, John Wiley & Sons, 1984. (Good for multiple-comparison
procedures.)
- Applied Linear Statistical Models, 4th ed., John Neter,
M. Kutner, C. J. Nachtsheim, and W. Wasserman, Irwin/McGraw Hill, 1999.
- Applied Multivariate Statistical Analysis. 5th ed., R.
A. Johnson and D. W. Wichern, Prentice Hall, 2002.