{"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/dqn-tamer-human-in-the-loop-reinforcement","title":"DQN-TAMER: Human-in-the-Loop Reinforcement Learning with Intractable Feedback","arxiv_id":"1810.11748","date":"2018-10-28","proceeding":null,"authors":["Riku Arakawa","Sosuke Kobayashi","Yuya Unno","Yuta Tsuboi","Shin-ichi Maeda"],"abstract":"Exploration has been one of the greatest challenges in reinforcement learning\n(RL), which is a large obstacle in the application of RL to robotics. Even with\nstate-of-the-art RL algorithms, building a well-learned agent often requires\ntoo many trials, mainly due to the difficulty of matching its actions with\nrewards in the distant future. A remedy for this is to train an agent with\nreal-time feedback from a human observer who immediately gives rewards for some\nactions. This study tackles a series of challenges for introducing such a\nhuman-in-the-loop RL scheme. The first contribution of this work is our\nexperiments with a precisely modeled human observer: binary, delay,\nstochasticity, unsustainability, and natural reaction. We also propose an RL\nmethod called DQN-TAMER, which efficiently uses both human feedback and distant\nrewards. We find that DQN-TAMER agents outperform their baselines in Maze and\nTaxi simulated environments. Furthermore, we demonstrate a real-world\nhuman-in-the-loop RL application where a camera automatically recognizes a\nuser's facial expressions as feedback to the agent while the agent explores a\nmaze.","url_abs":"http://arxiv.org/abs/1810.11748v1","url_pdf":"http://arxiv.org/pdf/1810.11748v1.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":"dqn-tamer-human-in-the-loop-reinforcement","repo_url":"https://github.com/JulienDesvergnes/human-reinforcement-learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"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":"https://syntology.ai/paper/1810.11748","atlas_url":"https://app.syntology.ai/?focus=1810.11748","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}