{"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/perseus-randomized-point-based-value","title":"Perseus: Randomized Point-based Value Iteration for POMDPs","arxiv_id":"1109.2145","date":"2011-09-09","proceeding":null,"authors":["M. T. J. Spaan","N. Vlassis"],"abstract":"Partially observable Markov decision processes (POMDPs) form an attractive\nand principled framework for agent planning under uncertainty. Point-based\napproximate techniques for POMDPs compute a policy based on a finite set of\npoints collected in advance from the agents belief space. We present a\nrandomized point-based value iteration algorithm called Perseus. The algorithm\nperforms approximate value backup stages, ensuring that in each backup stage\nthe value of each point in the belief set is improved; the key observation is\nthat a single backup may improve the value of many belief points. Contrary to\nother point-based methods, Perseus backs up only a (randomly selected) subset\nof points in the belief set, sufficient for improving the value of each belief\npoint in the set. We show how the same idea can be extended to dealing with\ncontinuous action spaces. Experimental results show the potential of Perseus in\nlarge scale POMDP problems.","url_abs":"http://arxiv.org/abs/1109.2145v1","url_pdf":"http://arxiv.org/pdf/1109.2145v1.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":"perseus-randomized-point-based-value","repo_url":"https://github.com/ilyasshaloui/research-topics","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1109.2145","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}