This repository provides an implementation for a soft action particle method (SAPM) for continuous action space problems of reinforcement learning. SAPM efficiently reduces the number of parameter of policy network while maintaining the performance. More detail algorithm can be found in our paper. We develope our code based on OpenAI's spinningup.
Minjae Kang*, Kyungjae Lee*, and Songhwai Oh, "Soft Action Particle Deep Reinforcement Learning for a Continuous Action Space," in Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Nov. 2019.
*: Equal contribution
Distributed under the MIT License. See LICENSE for more information.
- This work was supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No. 2019-0-01371, Development of Brain-Inspired AI with Human-Like Intelligence).