Posts by Collection

publications

Temporally Extended Successor Representations

Published in The Multi-disciplinary Conference on Reinforcement Learning and Decision Making, 2022

In this conference submission, I outline some preliminary results showing the efficacy of using successor representations that represent the expected state dynamics given an action has a finite horizon (a numver of action repeats), which is also generated under some policy. Link to Paper

talks

teaching

Mathematics I

Undergraduate course, Imperial College London, Department of Bioengineering, 2019