Academic Year 2023/24, Semester 1
Department of Mathematics, The University of Manchester

The course starts on the 26th September 2023 and runs in semester weeks W1-W5 and W7-W12. There is no teaching in W6.

Teaching staff:

Lecturer: Korbinian Strimmer
Office hour: Friday, 3pm (after the tutorial). Please send email for an appointment.

Graduate teaching assistants: tba

Overview and syllabus:

The MATH38161 module is an introductory course in Multivariate Statistics and Machine Learning for third year mathematics students. This module provides an introduction to classical multivariate statistics (multivariate random variables and distributions, transformations, dimension reduction methods such as PCA and CCA, multivariate dependence measures, multivariate mixture models) and also offers a primer to machine learning (unsupervised learning and clustering, supervised learning and classification, EM algorithm). At the end of this module there is a brief outlook to modern nonlinear and nonparametric machine learning methods (e.g. random forests, Gaussian processes, neural networks).

The modules is designed to run over the course of 11 weeks in six parts, each covering a particular aspect of multivariate statistics and machine learning:

  1. Multivariate random variables and estimation in large and small sample settings (W1 and W2)
  2. Transformations and dimension reduction (W3 and W4)
  3. Unsupervised learning/clustering (W5 and W7)
  4. Supervised learning/classification (W8 and W9)
  5. Measuring and modelling multivariate dependencies (W10)
  6. Nonlinear and nonparametric models (W11, W12)

The presentation of the material focuses on concepts and methods, not on theory. The practical implementation and application in R is another priority in this module. The example sheets contain both computational and programming problems as well as theoretical problems.

Teaching Format:

This course is taught in traditional format with 2 lectures and 1 tutorial per week.

The main reference in this course are the MATH38161 lecture notes and you are expected to revisit the content of the lectures each week using these notes. The weekly learning plan available on Blackboard indicates which parts of the notes are relevant in each study week.

In the weekly tutorials the problems in worksheets are discussed. You are expected to attempt the problems within the week before the tutorials. Often this will involve some bits of programming and data analysis on the computer. You are welcome to work together in a group to solve the problems.

The teaching style follows a traditional "non-flipped" format. Hence, you don't need to study the content before coming to the lectures. However, it is essential to revisit the material of each week before the next lectures and also solve the corresponding example sheet so that you don't fall behind and continue to be able to follow the subsequent lectures.

All lectures (but not the tutorials!) will be recorded automatically and corresponding podcasts will appear on the UoM video portal within 24 hours of the event.

Course materials:

See also the MATH38161 UoM library reading list

To get started, download the lecture notes:

In addition, download the weekly learning plan and worksheets with solutions from Blackboard.

Further information such as coursework instructions, previous exam papers, feedback to students etc. is available on Blackboard.

Dates and location:

The lectures and tutorials take place at the following dates and locations:

Session Location Day Time Semester 1 Week
Lecture 1 Zochonis TH B Tuesday 11:00 W1-W5, W7-W12
Lecture 2 Zochonis TH B Tuesday 12:00 W1-W5, W7-W12
Tutorial Alan Turing G205 Friday 14:00 W1-W5, W7-W12

Assessment:

There is one in-semester take-home assessment worth 20%. This is small project requiring data analysis in R and writing of a corresponding statistical report, preferably in R Markdown.

The end-of-semester assessment is worth the remaining 80% and is concerned with theory and methods. This exam will take place on campus in Manchester on 22 January 2023.

Assessment Date Semester 1 Week
Project work (20%): Announced: Monday 13 November 2023, 12 noon
Submission: Monday 27 November 2023, 12 noon		 Work on project in Semester W8 and W9
Exam (80%): Monday 22 January 2024, 2pm-4pm Exam period

The instructions for the in-semester project will be made available on Blackboard on Monday 13 November, 12 noon. The expected amount of time to complete the project is 10h.

Frequently asked questions and comments:

I hope you enjoy the course! If you have any questions, comments, or corrections please contact the lecturer. First check the MATH38161 FAQ whether your question has already been asked and answered before!