Riemannian Proximal Policy Optimization
- Shijun Wang
- Baocheng Zhu
- Chen Li
- Mingzhe Wu
- James Zhang
- Wei Chu
- Yuan Qi
Abstract
In this paper, we propose a general Riemannian proximal optimization algorithm with guaranteed convergence to solve Markov decision process (MDP) problems. To model policy functions in MDP, we employ Gaussian mixture model (GMM) and formulate it as a non-convex optimization problem in the Riemannian space of positive semidefinite matrices. For two given policy functions, we also provide its lower bound on policy improvement by using bounds derived from the Wasserstein distance of GMMs. Preliminary experiments show the efficacy of our proposed Riemannian proximal policy optimization algorithm.
- Full Text:
PDF
- DOI:10.5539/cis.v13n3p93
Journal Metrics
WJCI (2020): 0.439
Impact Factor 2020 (by WJCI): 0.247
Google Scholar Citations (March 2022): 6907
Google-based Impact Factor (2021): 0.68
h-index (December 2021): 37
i10-index (December 2021): 172
(Click Here to Learn More)
Index
- Academic Journals Database
- BASE (Bielefeld Academic Search Engine)
- CiteFactor
- CNKI Scholar
- COPAC
- CrossRef
- DBLP (2008-2019)
- EBSCOhost
- EuroPub Database
- Excellence in Research for Australia (ERA)
- Genamics JournalSeek
- Google Scholar
- HOLLIS (Harvard Library)
- Infotrieve
- LOCKSS
- Mendeley
- NewJour
- Open J-Gate
- PKP Open Archives Harvester
- Publons
- ResearchGate
- Scilit
- SHERPA/RoMEO
- Standard Periodical Directory
- The Index of Information Systems Journals
- The Keepers Registry
- UCR Library
- Universe Digital Library
- WJCI Report
- WorldCat
Contact
- Chris LeeEditorial Assistant
- cis@ccsenet.org