Can Deep Reinforcement Learning Solve Erdos-Selfridge-Spencer Games?
arXiv:1711.02301
Abstract
Deep reinforcement learning has achieved many recent successes, but our understanding of its strengths and limitations is hampered by the lack of rich environments in which we can fully characterize optimal behavior, and correspondingly diagnose individual actions against such a characterization. Here we consider a family of combinatorial games, arising from work of Erdos, Selfridge, and Spencer, and we propose their use as environments for evaluating and comparing different approaches to reinforcement learning. These games have a number of appealing features: they are challenging for current learning approaches, but they form (i) a low-dimensional, simply parametrized environment where (ii) there is a linear closed form solution for optimal behavior from any state, and (iii) the difficulty of the game can be tuned by changing environment parameters in an interpretable way. We use these Erdos-Selfridge-Spencer games not only to compare different algorithms, but test for generalization, make comparisons to supervised learning, analyse multiagent play, and even develop a self play algorithm. Code can be found at: https://github.com/rubai5/ESS_Game
Accepted to ICML 2018, code opensourced at: https://github.com/rubai5/ESS_Game
Cited by in corpus (7)
- A Survey and Critique of Multiagent Deep Reinforcement Learning
- A Study on Overfitting in Deep Reinforcement Learning
- Playing the lottery with rewards and multiple languages: lottery tickets in RL and NLP
- Off-Policy Evaluation via Off-Policy Classification
- Iroko: A Framework to Prototype Reinforcement Learning for Data Center Traffic Control
- Pointer Value Retrieval: A new benchmark for understanding the limits of neural network generalization
- Competitive Policy Optimization