Width-based Lookaheads with Learnt Base Policies and Heuristics Over the Atari-2600 Benchmark

Stefan O'Toole, Nir Lipovetzky, Miquel Ramirez, Adrian Pearce

Keywords: width based planning, planning and learning, online planning, atari, planning over simulators

Abstract: We propose new width-based planning and learning algorithms inspired from a careful analysis of the design decisions made by previous width-based planners. The algorithms are applied over the Atari-2600 games and our best performing algorithm, Novelty guided Critical Path Learning (N-CPL), outperforms the previously introduced width-based planning and learning algorithms $\pi$-IW(1), $\pi$-IW(1)+ and $\pi$-HIW(n, 1). Furthermore, we present a taxonomy of the Atari-2600 games according to some of their defining characteristics. This analysis of the games provides further insight into the behaviour and performance of the algorithms introduced. Namely, for games with large branching factors, and games with sparse meaningful rewards, N-CPL outperforms $\pi$-IW, $\pi$-IW(1)+ and $\pi$-HIW(n, 1).

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

TL;DR: We propose new width-based planning and learning algorithms which we apply over the Atari-2600 games.

Code: https://github.com/stefanotoole/N-CPL