MAS467 Linear and Generalised Linear Models
|Semester 1, 2019/20||10 Credits|
|Lecturer:||Dr Jonathan Jordan||uses MOLE||Reading List|
The module will further develop the general theory of linear models, building on theory taught in MAS223. Extensions from the L2 course will include methods for dealing with large numbers of independent variables. The module will also introduce generalised linear models, which can be used for modelling data such as binary data and count data, where a normal distribution would not be appropriate. These developments dramatically extend the range of problems that can be studied. The methods will be implemented using R.
Prerequisites: MAS223 (Statistical Inference and Modelling)
No other modules have this module as a prerequisite.
- Review of Linear Regression: Simple and multiple linear regression, linear model in matrix form, LS estimators and its distributions.
- General hypothesis testing: Distribution of LS estimators, confidence intervals, the general hypothesis testing, ANOVA tables.
- Transformations and variable selection: Variance stabilization transformations, Box-Cox transformations, variable selection, automated methods for variable selection.
- Introduction to generalized linear models: Introduction, definition, GLM distributions.
- GLM estimation: Parameter estimation, use of deviance in GLMs to test model fit. Model building (analysis of deviance), types of residuals, quasi likelihood.
- Binary response: Likelihood, links, examples, odds ratio, logistic regression.
- Poisson regression: Poisson regression for count data, using offsets to adjust for exposure.
- Two way contingency tables: Response & controlled variables, association and homogeneity, probability distributions for two-way tables.
- To review and extend the students knowledge of the standard linear model, building on concepts introduced at L2.
- To introduce the theory of generalised linear models.
- To show how these methods are applied to data, and what kinds of conclusions are possible.
- To demonstrate the fitting and interpretation of linear and generalised linear models to data using the statistical computing language R.
- Obtain a technical understanding and appreciation of ordinary and generalised linear modelling methods.
- Be able to identify circumstances in which ordinary and generalised linear models can be used for data analysis, and understand what conclusions and inferences can be drawn.
- Know how to fit linear and generalised linear models using R, and interpret the output
Lectures, problem solving
No lectures, no tutorials
One formal 2 hour open- book written examination [70%] Format: 3 compulsory questions. Project [30%].
|C||Atkinson||Plots, Transformations and Regression||519.51 (A)||Blackwells||Amazon|
|C||Dobson, A.J. and Adrian Barnett||An Introduction to Generalized Linear Models|
|C||Draper and Smith||Applied Regression Analysis||519.536 (D)||Blackwells||Amazon|
|C||McCullagh, P J and Nelder, J A||Generalised Linear Models||Blackwells||Amazon|
|C||Montgomery, Peck and Vining||Introduction to Linear Regression Analysis||519.51 (M)||Blackwells||Amazon|
|C||Seber and Lee||Linear Regression Analysis||519.51 (S)||Blackwells||Amazon|
(A = essential, B = recommended, C = background.)
Most books on reading lists should also be available from the Blackwells shop at Jessop West.