Maths education seminar: Jonas Latz an analogy between machine learning and a child learning through a game
| Dates: | 11 May 2026 |
|---|
| Times: | 14:00 - 15:00 |
|---|
| What is it: | Seminar |
|---|
| Organiser: | Department of Mathematics |
|---|
| Who is it for: | University staff |
|---|
| Speaker: | Jonas Latz |
|---|
|
Speaker: Jonas Latz (University of Manchester)
Title: Machine Learning is (not) child’s play
Abstract: Machine learning describes the generation of an artificial intelligence (AI) through data. The terminology creates an analogy between this process and the learning in intelligent beings: An analogy that has been criticised as inaccurate in the past. Goal of this note is to create a better analogy explaining machine learning – especially the mathematical optimisation process at its foundation. The analogy is given through a children’s game that is able to not only accurately describe the machine learning process, but to also show challenges arising within the process, and how these are solved within the state of the art in machine learning. This note acts as a brief introduction to optimisation in machine learning, it proposes and discusses the new analogy to the machine learning process, and it aims to inform teaching practice.
Preprint: https://latzplacian.org/wp-content/uploads/2025/12/Latz2025_MLisNotChildsPlay_v1_1.pdf
Speaker
Jonas Latz
Role: Lecturer in Applied Mathematics
Organisation: University of Manchester
Travel and Contact Information
Find event
G.114
Alan Turing Building
Manchester
