Module Catalogues, Xi'an Jiaotong-Liverpool University   
 
Module Code: MTH303
Module Title: Linear Statistical Models
Module Level: Level 3
Module Credits: 5.00
Academic Year: 2019/20
Semester: SEM1
Originating Department: Mathematical Sciences
Pre-requisites: N/A
   
Aims
To understand how regression methods for continuous data extend to include multiple continuous and categorical predictors and categorical response variables.


To gain a thorough understanding of the model diagnostics and building.


To develop familiarity with the computer package.
Learning outcomes 
A.Understand the rationale and assumptions of linear regression and analysis of variance.

B.Carry out and interpret linear regressions and analyses of variance, and derive appropriate theoretical results.


C. Know the commonly used techniques for model diagnostics and variable selection.


D. Understand the rationale and assumptions of generalized linear models.


E. Carry out and interpret analyses involving generalized linear models, and derive appropriate theoretical results.


F. Perform linear regression and generalized linear models analysis using computer package or its outputs.

Method of teaching and learning 
This module will be delivered by a combination of formal lectures and tutorials.
Syllabus 
-SIMPLE and MULTIPLE LINEAR REGRESSIONS


Model assumptions; Estimation and its inference; Adequacy of the model; Mean response and its confidence interval; Prediction and its inference; Test of hypotheses; Analysis of variance; Analysis of covariance


-MODEL DIAGNOSTICS and VARIABLE SELECTION


Residual analyses; Transformations to correct model inadequacies; Diagnostics for Leverage; Influence and outliers; Collinearity; Model selection; Model validation (Pearson’s chi-square test and the likelihood ratio test); Implementation of GLM using available computer packages and data sets.



-GENERALIZED LINEAR MODELS


Exponential family (typical examples; mean, variance, link function) and generalized linear models; Logistic regression; Nominal and ordinal Logistic regression; Poisson regression and Log linear model; Gamma regression
Delivery Hours  
Lectures Seminars Tutorials Lab/Prcaticals Fieldwork / Placement Other(Private study) Total
Hours/Semester 39     13        52 

Assessment

Sequence Method % of Final Mark
1 Formal Examination 85.00
2 Coursework 15.00

Module Catalogue generated from SITS CUT-OFF: 12/9/2019 11:51:19 PM