Academic Journal
Generalized Policy Improvement Algorithms with Theoretically Supported Sample Reuse.
| Τίτλος: | Generalized Policy Improvement Algorithms with Theoretically Supported Sample Reuse. |
|---|---|
| Συγγραφείς: | Queeney J; Mitsubishi Electric Research Laboratories, Cambridge, MA 02139 USA. He performed the majority of this work while with the Division of Systems Engineering, Boston University, Boston, MA 02215 USA., Paschalidis IC; Department of Electrical and Computer Engineering and Division of Systems Engineering, Boston University, Boston, MA 02215 USA., Cassandras CG; Department of Electrical and Computer Engineering and Division of Systems Engineering, Boston University, Boston, MA 02215 USA. |
| Πηγή: | IEEE transactions on automatic control [IEEE Trans Automat Contr] 2025 Feb; Vol. 70 (2), pp. 1236-1243. Date of Electronic Publication: 2024 Sep 03. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: Institute of Electrical and Electronics Engineers Country of Publication: United States NLM ID: 101210419 Publication Model: Print-Electronic Cited Medium: Print ISSN: 0018-9286 (Print) Linking ISSN: 00189286 NLM ISO Abbreviation: IEEE Trans Automat Contr Subsets: PubMed not MEDLINE |
| Imprint Name(s): | Original Publication: New York, N.Y. : Institute of Electrical and Electronics Engineers, 1963- |
| Περίληψη: | We develop a new class of model-free deep reinforcement learning algorithms for data-driven, learning-based control. Our Generalized Policy Improvement algorithms combine the policy improvement guarantees of on-policy methods with the efficiency of sample reuse, addressing a trade-off between two important deployment requirements for real-world control: (i) practical performance guarantees and (ii) data efficiency. We demonstrate the benefits of this new class of algorithms through extensive experimental analysis on a broad range of simulated control tasks. |
| References: | IEEE Trans Neural Netw Learn Syst. 2022 May;33(5):2223-2235. (PMID: 33481718) IEEE Trans Cybern. 2022 Sep;52(9):9428-9438. (PMID: 33705327) |
| Grant Information: | R01 GM093147 United States GM NIGMS NIH HHS; R01 GM135930 United States GM NIGMS NIH HHS; UL1 TR001430 United States TR NCATS NIH HHS |
| Contributed Indexing: | Keywords: Policy improvement; policy optimization; reinforcement learning; sample reuse |
| Entry Date(s): | Date Created: 20250820 Latest Revision: 20250904 |
| Update Code: | 20260130 |
| PubMed Central ID: | PMC12360665 |
| DOI: | 10.1109/tac.2024.3454011 |
| PMID: | 40832367 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 0018-9286 |
|---|---|
| DOI: | 10.1109/tac.2024.3454011 |