Skip to main content

Trinity College Dublin, The University of Dublin

Menu Search

Module Descriptor School of Computer Science and Statistics

Module CodeST3451
Semester TaughtMichaelmas
Contact Hours

3 hours per week, some of which will be tutorials

Module PersonnelRozenn Dahyot
Learning Outcomes

When students have successfully completed this module they should be able to:

  1. Derive and apply estimators, tests and confidence intervals for the parameters for a range of linear regression models.
  2. Derive and construct ANOVA tables.
  3. Examine the fit of a regression model through regression diagnostics and test the assumptions of the model.
  4. Build an appropriate linear regression model for a given data set.  
Learning Aims

The student will learn about the simple linear regression (SLR) model in detail. This will include derivation of least squares estimators and their properties, sampling distributions of the estimators in the case of Gaussian errors, and tests of significance. The student will also learn about ANOVA- decomposition of the error sum of squares. The matrix approach to linear regression will follow where multiple regression will be discussed. Various diagnostics of fit will be explored, with illustration of how these can be used in practice. Some modifications of the usual regression model will be discussed as well as model building through variable selection.

Module Content
  1. Simple linear regression
  2. Multiple regression
  3. Regression diagnostics
  4. Variable selection
Recommended Reading List
  1. A second course in statistics regression analysis- Mendenhall & Sincich
  2. Classical & modern regression with applications- R. H. Meyers
  3. Introduction to linear regression analysis- Montgomery, Peck & Vining Applied regression analysis- Draper & Smith 
Module Prerequisites
Assessment Details

100% Exam

Module Website
Academic Year of Data2017/18