STATISTICS 545 (2005-2006, Term 1)

Course Outline (REVISED VERSION OF SEPT. 6, pdf file)

Assigned coursework (will be added onto as we cover material).

Page last updated: Nov. 22, 2005

Lecture 1: Statistical Principles I: Estimation (handwritten notes).

Lecture 2: Statistical Principles II: Uncertainty assessment and hypothesis testing (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 3: Nice things about R (slides - pdf: fullsize, reduced, ps: fullsize, reduced ), and three example tasks (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 4: Linear models, Part I (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 5: Linear models, Part II: handwritten slides on regression diagnostics.

Lecture 6: Logistic regression (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 7: Generalized linear models I (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 8: Generalized linear models II - Overdispersion (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 9: Generalized linear models III - Log-linear modelling (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

We lapped ourselves. It took us 10 classes to cover lectures 1 through 9. Hence there is no lecture 10.

Lecture 11: The bootstrap (slides - pdf: fullsize, reduced, ps: fullsize, reduced ). Also, here is R code for examples one, two, and three.

Lecture 12: Model Choice (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 13: More Model Choice (slides - pdf: fullsize, reduced, ps: fullsize, reduced ). Also, R code for the stepwise and cross-validation examples.

Lapped ourselves again - 4 classes to cover lectures 11 through 13 - Hence there is no lecture 14.

Lecture 15: Nonlinear Regression (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 16: Expectation-Maximization (EM) Algorithm (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 17: Simulation Studies (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 18: Handwritten notes on hierarchical models (otherwise known as random-effect models, mixed models, or random coefficient models).

Lecture 19: More on hierarchical models, including the Sitka data ex.: (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 20: Curve-fitting. No pre-fab slides, but some pictures ( pdf, or ps), and a bit of code.

Lecture 21: Move from curve-fitting to additive models, with this example: (slides - pdf: fullsize, reduced, ps: fullsize, reduced ).

Lecture 22: We didn't really get started on additive models last time, so that discussion carrys over. We may also start talking about "tree models" if time permits.

Lecture 23: Tree models. Here are the examples. ( pdf: fullsize, reduced, ps: fullsize, reduced ).