{"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/lift-reinforcement-learning-in-computer","title":"LIFT: Reinforcement Learning in Computer Systems by Learning From Demonstrations","arxiv_id":"1808.07903","date":"2018-08-23","proceeding":null,"authors":["Michael Schaarschmidt","Alexander Kuhnle","Ben Ellis","Kai Fricke","Felix Gessert","Eiko Yoneki"],"abstract":"Reinforcement learning approaches have long appealed to the data management\ncommunity due to their ability to learn to control dynamic behavior from raw\nsystem performance. Recent successes in combining deep neural networks with\nreinforcement learning have sparked significant new interest in this domain.\nHowever, practical solutions remain elusive due to large training data\nrequirements, algorithmic instability, and lack of standard tools. In this\nwork, we introduce LIFT, an end-to-end software stack for applying deep\nreinforcement learning to data management tasks. While prior work has\nfrequently explored applications in simulations, LIFT centers on utilizing\nhuman expertise to learn from demonstrations, thus lowering online training\ntimes. We further introduce TensorForce, a TensorFlow library for applied deep\nreinforcement learning exposing a unified declarative interface to common RL\nalgorithms, thus providing a backend to LIFT. We demonstrate the utility of\nLIFT in two case studies in database compound indexing and resource management\nin stream processing. Results show LIFT controllers initialized from\ndemonstrations can outperform human baselines and heuristics across latency\nmetrics and space usage by up to 70%.","url_abs":"http://arxiv.org/abs/1808.07903v1","url_pdf":"http://arxiv.org/pdf/1808.07903v1.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":"lift-reinforcement-learning-in-computer","repo_url":"https://github.com/reinforceio/tensorforce","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"lift-reinforcement-learning-in-computer","repo_url":"https://github.com/CAVED123/Tensorforce","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"lift-reinforcement-learning-in-computer","repo_url":"https://github.com/CAVED123/tensorforce-","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"lift-reinforcement-learning-in-computer","repo_url":"https://github.com/tensorforce/tensorforce","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"task_slug":"management","task_name":"Management"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}