Reinforcement Learning: Straight Up and Upside Down
Presenter | 講演者
Researcher at Araya
Biography | 略歴
Kai is a researcher at Araya, where he works on deep learning, reinforcement learning, and neuroscience. He received his BA in Computer Science at the University of Cambridge in 2012 and his PhD in Bioengineering from Imperial College London in 2020. In the past he has worked at DeepMind, Microsoft Research, Facebook AI Research, Twitter and NNAISENSE, and was also a mentor for the OpenAI Scholars program.
Abstract | 概要
Reinforcement learning is the study of sequential decision making, with the “simple” objective of maximising long-term reward. In this talk I will introduce the two main approaches to solving reinforcement learning, evaluating their pros and cons. To finish, I will introduce upside down reinforcement learning, which may provide a simplified approach to tackling many traditional reinforcement learning problems.