551: Deep Reinforcement Learning — with Wah Loon Keng
Super Data Science: ML & AI Podcast with Jon Krohn
00:00
Evolution of Deep Reinforcement Learning Algorithms
The chapter traces the history and development of deep reinforcement learning algorithms, from the initial successes in backgammon and chess to the advancements made by DeepMind with algorithms like DQN in 2013. It discusses the application of neural networks in reinforcement learning, achievements in Atari games with double DQN and prioritize experience replay, and the use of proximal policy optimization (PPO) in Dota in 2017. The conversation also explores the evolution from AlphaGo to AlphaZero and AlphaStar, showcasing the progress made by DeepMind and OpenAI in mastering various games and applications.
Play episode from 18:01
Transcript
