Can Deep Reinforcement Learning Solve ErdosSelfridgeSpencer Games?
This topic contains 0 replies, has 1 voice, and was last updated by arXiv 1 year, 4 months ago.

Can Deep Reinforcement Learning Solve ErdosSelfridgeSpencer Games?
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 lowdimensional, 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 ErdosSelfridgeSpencer games not only to compare different algorithms, but also to compare approaches based on supervised and reinforcement learning, to analyze the power of multiagent approaches in improving performance, and to evaluate generalization to environments outside the training set.
Can Deep Reinforcement Learning Solve ErdosSelfridgeSpencer Games?
by Maithra Raghu, Alex Irpan, Jacob Andreas, Robert Kleinberg, Quoc V. Le, Jon Kleinberg
https://arxiv.org/pdf/1711.02301v2.pdf
You must be logged in to reply to this topic.