Preface
About the module
Topics covered
The MATH38161 module is designed to run over the course of 11 weeks. It has six parts, each covering a particular aspect of multivariate statistics and machine learning:
- Multivariate random variables and estimation in large and small sample settings (W1 and W2)
- Transformations and dimension reduction (W3 and W4)
- Unsupervised learning/clustering (W5 and W6)
- Supervised learning/classification (W7 and W8)
- Measuring and modelling multivariate dependencies (W9)
- Nonlinear and nonparametric models (W10, W11)
This module focuses on:
- Concepts and methods (not on theory)
- Implementation and application in R
- Practical data analysis and interpretation (incl. report writing)
- Modern tools in data science and statistics (R markdown, R studio)
Additional support material
If you are a University of Manchester student and enrolled in this module you will find on Blackboard:
- a weekly learning plan for an 11 week study period,
- weekly worksheets with with examples (theory and application in R) and solutions in R Markdown, and
- exam papers of previous years.
Furthermore, there is also an MATH38161 online reading list hosted by the University of Manchester library.
Prerequisites
Multivariate statistics relies heavily on matrix algebra and vector and matrix calculus. For a refresher of the essentials please refer to the supplementary
Furthermore, this module builds on earlier statistics modules, especially on likelihood estimation and Bayesian statistics as discussed, e.g., in the module
For an overview of essential probability distributions see the
Acknowledgements
Many thanks to Beatriz Costa Gomes for her help to compile the first draft of these course notes in the winter term 2018 while she was a graduate teaching assistant for this course. I also thank the many students who suggested corrections.