Computationally Intensive Methods

Module code: MD7443

This module will cover a variety of computationally intensive methods including maximum likelihood estimation, modelling of non-linear effects, simulation studies, dealing with missing data and machine learning models.

Topics covered

• An introduction to the frequentist and Bayesian analysis of a statistical model with several unknown parameters
• How to apply splines and fractional polynomials when fitting non-linear functions including how to choose between models and plot the fitted functions
• Missing data patterns, problems arising from missing data and the basic concepts of multiple imputation
• The use of simulation in statistics and in particular for assessing the performance of statistical methods in terms of bias and convergence. In addition the use of simulation for complex power calculations
• A practical introduction to range of machine learning methods and predictive modelling
• The use of cross-validation for testing modelling performance
• The use of clustering and dimensional reduction methods for large data sets.