One Computer Science MSc thesis to be presented on 10 April

Thursday, 10 April there will be a master thesis presentation in Computer Science at Lund University, Faculty of Engineering.

The presentation will take place only in Zoom (see link below).

Note to potential opponents: Register as an opponent to the presentation of your choice by sending an email to the examiner for that presentation (firstname.lastname@cs.lth.se). Do not forget to specify the presentation you register for! Note that the number of opponents may be limited (often to two), so you might be forced to choose another presentation if you register too late. Registrations are individual, just as the oppositions are! More instructions are found on this page.

14:15-15:00 in E:4130 (Lucas)

Presenters: Christopher Meltzer, Morteza Rezaei 
Title: A Comparative Analysis of Reinforcement Learning and Transformer Models in the context of Othello
Examiner: Jacek Malec
Supervisor: Volker Krueger (LTH)

In this thesis, we make a comparative analysis between Reinforcement Learning and Transformer Models. We conduct multiple different experiments to evaluate the benefits of utilizing the two different methods, we try to train the Transformer using Reinforcement Learning techniques, and we experiment with the effect of fine-tuning the Transformer.

We found that the Reinforcement Learning in most cases performed better than the Transformer implementation. For the Transformer to be able to perform better than the Reinforcement Learning model, it needed to first be trained on a large synthetic dataset, and then be fine-tuned on a good dataset containing championship matches. In neither of the experiments where we trained the Transformer using Reinforcement Learning techniques did the Transformer learn anything useful.

Link to popular science summary: https://fileadmin.cs.lth.se/cs/Education/Examensarbete/Popsci/250410_14MeltzerRezaei.pdf