PPO Dash: Improving Generalization in Deep Reinforcement Learning

Joe Booth

PPO Dash: Improving Generalization in Deep Reinforcement Learning

Machine Learning 2019-07-29 v3 Artificial Intelligence

Authors: Joe Booth

Abstract

Deep reinforcement learning is prone to overfitting, and traditional benchmarks such as Atari 2600 benchmark can exacerbate this problem. The Obstacle Tower Challenge addresses this by using randomized environments and separate seeds for training, validation, and test runs. This paper examines various improvements and best practices to the PPO algorithm using the Obstacle Tower Challenge to empirically study their impact with regards to generalization. Our experiments show that the combination provides state-of-the-art performance on the Obstacle Tower Challenge.

Keywords

hyperparameter optimization reinforcement learning policy gradient

Cite

@article{arxiv.1907.06704,
  title  = {PPO Dash: Improving Generalization in Deep Reinforcement Learning},
  author = {Joe Booth},
  journal= {arXiv preprint arXiv:1907.06704},
  year   = {2019}
}

Comments

8 pages, 3 figures

Related papers

View all related →