{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/stochastic-variance-reduced-policy-gradient","title":"Stochastic Variance-Reduced Policy Gradient","arxiv_id":"1806.05618","date":"2018-06-14","proceeding":"ICML 2018 7","authors":["Matteo Papini","Damiano Binaghi","Giuseppe Canonaco","Matteo Pirotta","Marcello Restelli"],"abstract":"In this paper, we propose a novel reinforcement- learning algorithm\nconsisting in a stochastic variance-reduced version of policy gradient for\nsolving Markov Decision Processes (MDPs). Stochastic variance-reduced gradient\n(SVRG) methods have proven to be very successful in supervised learning.\nHowever, their adaptation to policy gradient is not straightforward and needs\nto account for I) a non-concave objective func- tion; II) approximations in the\nfull gradient com- putation; and III) a non-stationary sampling pro- cess. The\nresult is SVRPG, a stochastic variance- reduced policy gradient algorithm that\nleverages on importance weights to preserve the unbiased- ness of the gradient\nestimate. Under standard as- sumptions on the MDP, we provide convergence\nguarantees for SVRPG with a convergence rate that is linear under increasing\nbatch sizes. Finally, we suggest practical variants of SVRPG, and we\nempirically evaluate them on continuous MDPs.","url_abs":"http://arxiv.org/abs/1806.05618v1","url_pdf":"http://arxiv.org/pdf/1806.05618v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"stochastic-variance-reduced-policy-gradient","repo_url":"https://github.com/Dam930/rllab","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.05618","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}