Manipulating the Distributions of Experience used for Self-Play Learning in Expert Iteration
This work addresses incremental improvements in self-play learning for game-playing AI, specifically for researchers and practitioners in reinforcement learning and board game domains.
The paper tackled the problem of improving training performance in Expert Iteration by manipulating the distribution of self-play data, using methods like weighting samples by episode duration, applying Prioritized Experience Replay, and using an exploratory policy. The result showed major improvements in early training for some games and minor average improvements across fourteen board games.
Expert Iteration (ExIt) is an effective framework for learning game-playing policies from self-play. ExIt involves training a policy to mimic the search behaviour of a tree search algorithm - such as Monte-Carlo tree search - and using the trained policy to guide it. The policy and the tree search can then iteratively improve each other, through experience gathered in self-play between instances of the guided tree search algorithm. This paper outlines three different approaches for manipulating the distribution of data collected from self-play, and the procedure that samples batches for learning updates from the collected data. Firstly, samples in batches are weighted based on the durations of the episodes in which they were originally experienced. Secondly, Prioritized Experience Replay is applied within the ExIt framework, to prioritise sampling experience from which we expect to obtain valuable training signals. Thirdly, a trained exploratory policy is used to diversify the trajectories experienced in self-play. This paper summarises the effects of these manipulations on training performance evaluated in fourteen different board games. We find major improvements in early training performance in some games, and minor improvements averaged over fourteen games.