arXiv (Cornell University) · 2020 · 137 citations · 14 references
ComputingComputer ScienceAlgorithmsArtificial IntelligenceMachine LearningDeep LearningGradient-based OptimizationNeural NetworksPolicy LearningReinforcement LearningReinforcement Learning FoundationsDeep Reinforcement LearningPolicy GradientsPolicy OptimizationProximal AlgorithmsProximal Policy OptimizationTrust Region Policy Optimization
We study the roots of algorithmic progress in deep policy gradient algorithms through a case study on two popular algorithms: Proximal Policy Optimization (PPO) and Trust Region Policy Optimization (TRPO). Specifically, we investigate the consequences of "code-level optimizations:" algorithm augmentations found only in implementations or described as auxiliary details to the core algorithm. Seemingly of secondary importance, such optimizations turn out to have a major impact on agent behavior. Our results show that they (a) are responsible for most of PPO's gain in cumulative reward over TRPO, and (b) fundamentally change how RL methods function. These insights show the difficulty and importance of attributing performance gains in deep reinforcement learning. Code for reproducing our results is available at https://github.com/MadryLab/implementation-matters .
14
Mastering the game of Go without human knowledge
David Silver, Julian Schrittwieser, Karen Simonyan et al. · Nature · 2017 · 9.2K citations
Engineering, Educational Psychology, Cognitive Psychology +7
Playing Atari with Deep Reinforcement Learning
Volodymyr Mnih, Koray Kavukcuoglu, David Silver et al. · arXiv (Cornell University) · 2013 · 5.1K citations · Full text