1. Grandmaster level in StarCraft II using multi-agent reinforcement learning
- Author
-
Tom Schaul, David Silver, James Molloy, Junhyuk Oh, Katrina McKinney, Oriol Vinyals, David H. Choi, Junyoung Chung, Tobias Pohlen, Dani Yogatama, Tobias Pfaff, Demis Hassabis, Michael Mathieu, Dan Horgan, Ivo Danihelka, Igor Babuschkin, Dario Wünsch, Tom Le Paine, Yury Sulsky, Wojciech Marian Czarnecki, Rémi Leblond, Ziyu Wang, Andrew Dudzik, Trevor Cai, Chris Apps, Yuhuai Wu, David Budden, Valentin Dalibard, Timo Ewalds, Oliver Smith, John P. Agapiou, Aja Huang, Roman Ring, Petko Georgiev, Max Jaderberg, Koray Kavukcuoglu, Alexander Vezhnevets, Caglar Gulcehre, Manuel Kroiss, Laurent Sifre, Richard E. Powell, and Timothy P. Lillicrap
- Subjects
Matching (statistics) ,Multidisciplinary ,Computer science ,ComputingMilieux_PERSONALCOMPUTING ,02 engineering and technology ,010501 environmental sciences ,League ,01 natural sciences ,Domain (software engineering) ,Video Games ,Artificial Intelligence ,Human–computer interaction ,Stepping stone ,0202 electrical engineering, electronic engineering, information engineering ,Humans ,Learning ,Reinforcement learning ,Learning methods ,020201 artificial intelligence & image processing ,Relevance (information retrieval) ,Reinforcement learning algorithm ,Reinforcement, Psychology ,0105 earth and related environmental sciences - Abstract
Many real-world applications require artificial agents to compete and coordinate with other agents in complex environments. As a stepping stone to this goal, the domain of StarCraft has emerged as an important challenge for artificial intelligence research, owing to its iconic and enduring status among the most difficult professional esports and its relevance to the real world in terms of its raw complexity and multi-agent challenges. Over the course of a decade and numerous competitions1-3, the strongest agents have simplified important aspects of the game, utilized superhuman capabilities, or employed hand-crafted sub-systems4. Despite these advantages, no previous agent has come close to matching the overall skill of top StarCraft players. We chose to address the challenge of StarCraft using general-purpose learning methods that are in principle applicable to other complex domains: a multi-agent reinforcement learning algorithm that uses data from both human and agent games within a diverse league of continually adapting strategies and counter-strategies, each represented by deep neural networks5,6. We evaluated our agent, AlphaStar, in the full game of StarCraft II, through a series of online games against human players. AlphaStar was rated at Grandmaster level for all three StarCraft races and above 99.8% of officially ranked human players.
- Published
- 2019