Reinforcement Learning: An Introduction
A textbook on how agents learn to make decisions through trial, error and rewards.
2018
103,831
Works
Selected research papers and books, linked to their authors.
Proximal Policy Optimization Algorithms
PPO improves an agent's behavior through repeated updates guided by trial and reward.
2017
48,053
Human-level control through deep reinforcement learning
Presented DQN, learning to play 49 Atari games from screen images and game scores.
2015
45,206
Mastering the game of Go with deep neural networks and tree search
AlphaGo combined neural networks and search to beat a professional Go player.
2016
26,603
Playing Atari with Deep Reinforcement Learning
Showed a neural network could learn Atari games directly from screen pixels and rewards.
2013
21,837
Deep Reinforcement Learning with Double Q-Learning
Double DQN reduces inflated reward estimates when learning through trial and error.
2016
14,524
Mastering the game of Go without human knowledge
AlphaGo Zero learned Go through self-play without examples of human games.
2017
14,508
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
AlphaZero learned chess, shogi and Go through self-play rather than human examples.
2018
10,329
Deep reinforcement learning from human preferences
Taught AI agents tasks using people's comparisons of their behavior rather than hand-written rewards.
2017
9,013
Grandmaster level in StarCraft II using multi-agent reinforcement learning
AlphaStar reached Grandmaster level in StarCraft II through competitive AI training.
2019
6,874