Geometry of Statistical Models

2026-2027

General Information

Course Code XM_0204
Credits 6 EC
Period P4+5
Course Level 400
Language of Tuition English
Faculty Faculty of Science
Course Coordinator dr. R. Hindriks
Examiner dr. R. Hindriks
Teaching Staff dr. R. Hindriks
Part of programme(s)

Practical Information

You need to register for this course yourself. The (de)registration deadlines can be found on VU.nl

Teaching Methods Seminar, Lecture

Course Objective

The aim of this master seminar is to learn how to read scientific texts at a research level, create and organise material, to give seminar talks to your fellow students, and be able to explain the mathematical concepts and ideas involved. It is also a goal to learn how to constructively participate in a research seminar by asking questions and giving feedback.

Course Content

The aim of this master seminar is to learn how to read scientific texts at a research level, create and organise material, to give seminar talks to your fellow students, and to be able to explain the mathematical concepts and ideas involved. It is also a goal to learn how to constructively participate in a research seminar by asking questions and giving feedback.

In this Master seminar we focus on the analysis of statistical models using methods from differential geometry. It is an active field with applications in diverse fields, including machine learing, neuroscience, and climatology. Specific learning goals are:

Understand how statistical models can be viewed as geometric objects using the Fisher information metric and Riemannian geometry;

Explain key ideas in information geometry, including statistical manifolds, geodesics, curvature, and dual flatness;

Apply core tools of Riemannian geometry such as tangent spaces, metrics, affine connections, parallel transport, and curvature to statistical models;

Perform explicit geometric calculations for common distributions, including normal, categorical, multinomial, and gamma families;

Describe the relationship between geometric concepts and statistical inference methods such as maximum likelihood estimation, estimator efficiency, and hypothesis testing;

Analyse the role of exponential and mixture connections in information geometry;

Explore applications of information geometry in machine learning, optimization, neural networks, and scientific modelling.

Additional Information Teaching Methods

A seminar, for two hours per week. At the end of the course, students choose a research article from a curated list, give a presentation, and write a report.

Method of Assessment

• Presentation 1 – 35%: General theory, based on the review article.

• Presentation 2 – 35%: Chosen topic, based on a selected research article.

• Oral exam – 30%.

• Peer feedback (pass/fail): Each student is also required to provide feedback on their peers’ presentations.

• Resit: There is no resit, due to the nature of the seminar.


For preparation, GenAI use is allowed at level 2b, GenAI assisted editing: GenAI content is allowed in the final submission but students must include a declaration of use of GenAI in their work stating how they used GenAI. During the assessment itself, there will be no access to GenAI, i.e., level 1 of GenAI use.

Entry Requirements

A seminar, for two hours per week. At the end of the course, students choose a research article from a curated list, give a presentation, and write a report.

Literature

Differential Geometry and Statistics, Michael K. Murray and John W. Rice, Chapman & Hall/CRC Routledge, 1993.

Additional notes provided during the course.

Research articles

Additional Information Target Audience

The course is designed to be accessible to all students (no specific background is required).

Additional Information

During the assessment, there will be no access to GenAI, i.e., level 1 of GenAI use.

Recommended background knowledge

Assumed is basic knowledge in linear algebra, calculus, probability theory, and statistics.