STU34501 – Applied Linear Statistical Methods I

Offered in 2023/24

Module CodeSTU34501
Module Name Applied Linear Statistical Methods I
ECTS Weighting [1]5 ECTS
Semester taughtSemester 1
Module Coordinator/s Dr. Jason Wyse

Module Learning Outcomes

On successful completion of this module, students will be able to:

LO1. Derive least squares estimators for a linear regression model

LO2. Derive and use properties of least squares estimators for inference

LO3. Extend the linear model to the general linear model (one way classification, polynomial regression) including use of dummy variables

LO4. Carry out model diagnostics through analysis of residuals

LO5. Form a Bayesian linear model and appreciate connections with ridge regression

LO6. Demonstrate how regularization can be used for model determination through the LASSO

Module Content

Working with linear and generalized linear models is an essential part of a data analyst’s work. This module presents the theory of the normal linear model and links this with the use of this theory in practice through examples in R. Diagnosing the fit (and hence appropriateness) of a model through residual analysis is discussed. The final part of the module looks at the more modern topic of regularization. This is motivated first through looking at the Bayesian linear model and its connections with ridge regression, then model determination through the Least Absolute Shrinkage and Selection Operator (LASSO) is discussed.

Teaching and learning Methods

There will be three classes per week. Some of these classes will be used for code demonstrations and tutorials.

Assessment Details

Assessment ComponentBrief Description Learning Outcomes Addressed% of totalWeek setWeek Due
ExaminationEnd of semester exam (2 hours)LO1-LO690%N/AN/A
AssignmentsFour assignments throughout the semesterLO1-LO610%3,5,7,94,6,8,10

Reassessment Details

 100% supplemental exam (2 hours)

Contact Hours and Indicative Student Workload

Contact Hours (scheduled hours per student over full module), broken down by: 33 hours
Lecture33
Independent study (outside scheduled contact hours), broken down by:82 hours
Preparation for classes and review of material (including preparation for examination, if applicable42
completion of assessments (including examination, if applicable)40
Total Hours115 hours

Recommended Reading List

Applied Linear Statistical Models, Michael Kutner, Christopher Nachtsheim, John Neter and William Li, McGraw-Hill/Irwin

Pattern Recognition and Machine Learning, Christopher Bishop, Springer

Computer Age Statistical Inference, Algorithms, Evidence and Data Science, Bradley Efron and Trevor Hastie, Cambridge University Press

Module Pre-requisites

STU23501

Module Co-requisites

None

Module Website

Blackboard