Projects per year
Abstract
Proximal Policy Optimization (PPO) is a highly popular model-free reinforcement learning (RL) approach. However, we observe that in a continuous action space, PPO can prematurely shrink the exploration variance, which leads to slow progress and may make the algorithm prone to getting stuck in local optima. Drawing inspiration from CMA-ES, a black-box evolutionary optimization method designed for robustness in similar situations, we propose PPO-CMA, a proximal policy optimization approach that adaptively expands the exploration variance to speed up progress. With only minor changes to PPO, our algorithm considerably improves performance in Roboschool continuous control benchmarks. Our results also show that PPO-CMA, as opposed to PPO, is significantly less sensitive to the choice of hyperparameters, allowing one to use it in complex movement optimization tasks without requiring tedious tuning.
Original language | English |
---|---|
Title of host publication | Proceedings of the 2020 IEEE 30th International Workshop on Machine Learning for Signal Processing, MLSP 2020 |
Publisher | IEEE |
Number of pages | 6 |
ISBN (Electronic) | 978-1-7281-6662-9 |
DOIs | |
Publication status | Published - Sept 2020 |
MoE publication type | A4 Conference publication |
Event | IEEE International Workshop on Machine Learning for Signal Processing - Aalto University, Espoo, Finland Duration: 21 Sept 2020 → 24 Sept 2020 Conference number: 30 https://ieeemlsp.cc |
Workshop
Workshop | IEEE International Workshop on Machine Learning for Signal Processing |
---|---|
Abbreviated title | MLSP |
Country/Territory | Finland |
City | Espoo |
Period | 21/09/2020 → 24/09/2020 |
Internet address |
Fingerprint
Dive into the research topics of 'PPO-CMA: Proximal Policy Optimization with Covariance Matrix Adaptation'. Together they form a unique fingerprint.Projects
- 2 Finished
-
Virtual Coach Based on Multibody Dynamics
Hämäläinen, P. (Principal investigator), Rajamäki, J. (Project Member), Naderi, K. (Project Member), Takatalo, J. (Project Member) & Kaos, M. (Project Member)
01/01/2017 → 31/12/2018
Project: Academy of Finland: Other research funding
-
IMAI: Interactive Movement Artificial Intelligence (IMAI)
Hämäläinen, P. (Principal investigator), Babadi, A. (Project Member), Rajamäki, J. (Project Member), Kaos, M. (Project Member), Takatalo, J. (Project Member), Toikka, J. (Project Member), Ikkala, A. (Project Member) & Naderi, K. (Project Member)
01/09/2016 → 31/08/2020
Project: Academy of Finland: Other research funding