Department:Mathematics and Statistics
Level:Part II (yr 3)
Course Convenor:Dr DM Berridge
back to top
- 70% Exam
- 10% Coursework
- 20% Project
Curriculum Design: Outline Syllabusback to top
Normal linear models: normal distribution, effects of covariates, likelihood inference
Logistic regression: binary data, Binomial distribution
Log-linear models and Poisson regression: count data, Poisson distribution
Generalized linear models (GLMs): exponential family distributions, the linear predictor, link function, likelihood inferenc
Model selection based on deviance.
The iteratively reweighted least squares (IRWLS) method.
Application of data sets.
* The R language is required for the laboratory sessions of MATH333. Combined major or minor students who do not take the whole of MATH390 should, if at all possible, attend the practical course in the R language which is given as part of MATH390, usually in week 6 or 7 of the summer term of their second year.
Single major mathematics students are required to write the project report using LaTeX, and so are combined major students who have taken the whole of MATH390. Other students are encouraged to audit the LaTeX course in MATH390 to enable them to use this to write their project, but they are permitted to use any other suitable word processor.
Curriculum Design: Pre-requisites/Co-requisites/Exclusionsback to top
MATH390 to be audited
Educational Aims: Subject Specific: Knowledge, Understanding and Skillsback to top
To understand the theoretical basis of generalized linear
models and to apply to a diverse range of practical problems.
To understand the effect of censoring in the statistical
analyses and to use appropriate statistical techniques for
Educational Aims: General: Knowledge, Understanding and Skillsback to top
To relate modern statistical models and methods to real life
situations and use relevant computer software for statistical
Learning Outcomes: Subject Specific: Knowledge, Understanding and Skillsback to top
At the end of the course the students should be able to demonstrate subject specific knowledge,
understanding and skills and have the ability to:
*Conduct exploratory data analyses using R for substantive applications
*Identify exponential family models and understand the IRWS algorithm
*Identify lifetime data and describe the effect of censoring in statistical inference
estimates for the survivor function
*Formulate sensible parametric models for a set of data, derive the likelihood functions and carry
out statistical inference
*Assess model fit and draw conclusions in nontechnical
Learning Outcomes: General: Knowledge, Understanding and Skillsback to top
On successful completion of this module the student should be able to
*Understand and explain the role of statistical models
*Justify and critique the use of statistical methods for reallife
Assessment: Details of AssessmentAssessment: Details of Assessmentback to top
Assessment will be through
(i) weekly coursework, aimed at testing and consolidating understanding of the basic
elements of the course;
(ii) a short project to test application of ideas and methods;
(iii) an examination in the Summer which assesses more fully the students' understanding and summative knowledge of the topics.
Curriculum Design: Single, Combined or Consortial Schemes to which the Module Contributesback to top
All MSci/BSc/BA single and combined major
degrees in Mathematics and Statistics.