STAT 846 (04) Computational Techniques in Statistics, 2007W
Longhai Li, Department of Mathematics and
Statistics, University of Saskatchewan
Description
This is a graduate course about computational techniques used in statistical inference. Topics will include computer arithmetic, the use of simulation to investigate the properties of statistical methods and to evaluate statistical methods, matrix computations used to implement linear models, such as QR and SVD descompositions, optimization methods used for maximum likelihood estimation, such as Newton-Raphson and EM algorithms, and various numerical integration methods for Bayesian inference, including a brief introduction to Markov chain Monte Carlo methods. The focus of this course is the computational aspects, but relevant statistical concepts will be explained as detailed as necessary. Assignments will involve writing R functions, which may call C functions for speed purpose, for some simple statistical problems not handled by standard packages. The R language will be briefly introduced at the start of the course.
Contact of instructor
Longhai Li, email:longhai[at]math.usask.ca, phone: 966-6095, office: MCLN
219.
Note: no lab and teaching assistant for this course.
Textbook
No textbook is required. The following books are recommended:
Lectures
Time: Thursday 9:30-12:30, from 10/01/2008-03/04/2008, except 21/02/2008
(reading week). 12 lectures in total.
Place: MCLN 242.2: Map
Office hours
Friday 4-5
Evaluation
FOUR assignments, each worth 15%.
TWO 1.5-hour mid-term tests, each worth 20%.
Course webpage
http://math.usask.ca/~longhai/teaching/stat846.s08
Lecture notes (PDF)
Click here to download.
Assignments (PDF)
Tests
Hand-outs
R Manuals
Click here for R
manuals.
Back to Longhai's homepage