Statistical Computing: Fall 2010

Project submissions (updated 11 November).

Some references:

• Dempster, Laird, and Rubin (1977). Maximum likelihood from incomplete data via the EM algorithm. [ JSTOR ]
• Redner and Walker (1984). Mixture densities, maximum likelihood and the EM algorithm. [ JSTOR ]

Projects (updated 10 October).

EM algorithm: Code samples discussed in class:

• Bivariate Normal with missing data
• ABO blood groups example

R demos:

• Kernel denisty estimation (needs Qt/qtbase)
• Nearest neighbour classification
• Nonparametric regression
• Importance sampling (output)

Code used to illustrate object-oriented programming, using Linear Congruential Generators as an example.

Summary of R code discussed in class (updated Friday August 27).

Summary of C code discussed in class (updated Friday August 6).

No classes August 9 - August 13

No classes July 21 - July 26

R tutorials. As part of your assignment for the first week, you should work through these tutorials, and solve the exercises in the "Language Overview II" and "Introduction to Statistical Inference" tutorials.

Assignment 1 (19 July, 2010)

Course Information

• Instructor: Deepayan Sarkar <deepayan.sarkar@gmail.com>
• Schedule: Mon Wed Fri 11:45 - 12:45, Fri 14:00 - 15:00 (Tutorial); July 19 - November 5, 2010 (Room 23).
• Office hours: Tue 14:00 - 15:00, Room 213, or by appointment

Outline (tentative)

• Computational Statistical Methods
• EM algorithm
• Simulation
• Resampling tecniques: jack-knife and bootstrap
• MCMC
• Programming
• Basic ideas of object-oriented programming
• Applications using R, C++, Python, (Javascript?)

Grading

• Assignments: 30%
• Mid-term Exam: 20%
• End-term Exam: 50%

Some assignments are to be done individually, others are group assignments. Grading will be based on classroom presentations and participation in classroom discussions of the assignments.