University of Galway

Course Module Information

Course Modules

Semester 2 | Credits: 5

An overview of modern approaches to statistical practice. Topics include: mixture models, nonparametric regression and smoothing, linear mixed models, penalised regression, classification and regression trees, prediction modelling, dealing with missing data, survival analysis
(Language of instruction: English)

Learning Outcomes
  1. Recognise how mixtures arise in an applied setting and construct a basic Gaussian mixture model
  2. Describe nonlinear relationships between variables using spline and kernel smoothers
  3. Extend linear regression to the case of clustered data, explain the concept of a random effect and its distinction from a fixed effect, calculate an intra-class correlation coefficient
  4. Apply penalised regression methods such as the LASSO for variable selection
  5. Construct a classification/regression tree; apply trees for prediction
  6. Calculate and interpret sensitivity, specificity, positive predictive value, negative predictive value; draw and interpret a ROC curve
  7. Derive and apply suitable methods for missing data
  8. Apply Cox proportional hazards models for time to event data
Assessments
  • Department-based Assessment (100%)
Teachers
The above information outlines module ST4040: "Modern Statistical Methods" and is valid from 2021 onwards.
Note: Module offerings and details may be subject to change.