Reinforcement Learning Reading Club (2021-2022)

I had the opportunity to participate in the Reinforcement Learning Reading Club, organized by Professor Edward Chung and hosted by my good friend Kaleb Ben Naveed, which brought together a group of interested researchers to study and share relevant papers.

Throughout the year, we went through fundamental RL algorithms, such as DQN, PPO and A3C, and also application-centric papers shared by postgraduate students who are applying RL in their own fields. I presented on two occasions, discussing the fundamental principles of Deep Q Networks and Deep Deterministic Policy Gradients.

The reading club has been an invaluable support on my learning journey. Learning RL alone challenging, even for people with a background in deep learning or computer vision, as RL is built upon concepts such as Temporal Difference Learning and Policy Gradients, which differ significantly from supervised deep learning methods. My personal learning (mostly from Sergey Levine and Pieter Abbeel’s online videos), followed by discussions with members in the reading club have given me essential knowledge that greatly benefits my work today.