{"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/gaussian-processes-for-data-efficient","title":"Gaussian Processes for Data-Efficient Learning in Robotics and Control","arxiv_id":"1502.02860","date":"2015-02-10","proceeding":null,"authors":["Marc Peter Deisenroth","Dieter Fox","Carl Edward Rasmussen"],"abstract":"Autonomous learning has been a promising direction in control and robotics\nfor more than a decade since data-driven learning allows to reduce the amount\nof engineering knowledge, which is otherwise required. However, autonomous\nreinforcement learning (RL) approaches typically require many interactions with\nthe system to learn controllers, which is a practical limitation in real\nsystems, such as robots, where many interactions can be impractical and time\nconsuming. To address this problem, current learning approaches typically\nrequire task-specific knowledge in form of expert demonstrations, realistic\nsimulators, pre-shaped policies, or specific knowledge about the underlying\ndynamics. In this article, we follow a different approach and speed up learning\nby extracting more information from data. In particular, we learn a\nprobabilistic, non-parametric Gaussian process transition model of the system.\nBy explicitly incorporating model uncertainty into long-term planning and\ncontroller learning our approach reduces the effects of model errors, a key\nproblem in model-based learning. Compared to state-of-the art RL our\nmodel-based policy search method achieves an unprecedented speed of learning.\nWe demonstrate its applicability to autonomous learning in real robot and\ncontrol tasks.","url_abs":"http://arxiv.org/abs/1502.02860v2","url_pdf":"http://arxiv.org/pdf/1502.02860v2.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":"gaussian-processes-for-data-efficient","repo_url":"https://github.com/mpd37/pilco-matlab","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"BSD-3-Clause"}}],"tasks":[{"task_slug":"gaussian-processes","task_name":"Gaussian Processes"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"}],"methods":[{"method_slug":"gaussian-process","method_name":"Gaussian Process"},{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1502.02860","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}