NEW

# Advanced Regression Analysis

Course category:
Masters
Module code:
MAT00042M
Year:
2012/13
Term:
Autumn
Credits:
20

Please note that this module is not available to those who have taken or are taking Generalised Linear Models

## Aims

This module is to teach students how to derive, from first principles and using matrix algebra, theoretical results relating to fitting regression models by least squares, local least squares or maximum likelihood approach, how to select a regression model to fit a given data set and carry out related statistical inferences using appropriate computer software.

## Learning objectives

At the end of the module you should

• have a reasonable ability to derive theoretical results relating to fitting regression models.

• have a reasonable ability to fit regression models to data, and carry out related statistical inferences using appropriate computer software.

• have a reasonable ability to use residual plots and other techniques to check the assumptions underlying regression analysis.

• have a reasonable ability to choose between alternative models for sets of data.

## Syllabus

• Exponential family and generalised linear models; Estimation (ML) and inference; Model selection; Model checking;
• Nonparametric regression models; Local linear modelling; Local least square estimation; Bandwidth selection; Varying-coefficient models.

## Recommended texts

N. R. Draper and H. Smith, Applied Regression Analysis, Wiley (1966, 1981, 1998)

S. Chatterjee and B. Price, Regression Analysis by Example, Wiley (1977, 1991, 1999).

P. McCullagh, J . Nelder, Generalized Linear Models, Second Edition. Boca Raton: Chapman and Hall/CRC (1989).

Fan, J. and Gijbels, I. Local Polynomial Modelling and its Applications (341pp). Chapman and Hall, London (1996).

## Teaching

• Autumn Term
• 3 lectures per week
• 1 computer practical or problems class per week

## Assessment

Two hours closed examination week 1 Spring Term 90%
Coursework 10%.  Note that coursework submitted after the advertised deadlines will be given a mark of zero.

## Elective information

This module is not available as an elective.

## Prerequisites

• Statistics , Linear Models