{"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/pre-training-with-non-expert-human","title":"Pre-training with Non-expert Human Demonstration for Deep Reinforcement Learning","arxiv_id":"1812.08904","date":"2018-12-21","proceeding":null,"authors":["Gabriel V. de la Cruz","Yunshu Du","Matthew E. Taylor"],"abstract":"Deep reinforcement learning (deep RL) has achieved superior performance in\ncomplex sequential tasks by using deep neural networks as function\napproximators to learn directly from raw input images. However, learning\ndirectly from raw images is data inefficient. The agent must learn feature\nrepresentation of complex states in addition to learning a policy. As a result,\ndeep RL typically suffers from slow learning speeds and often requires a\nprohibitively large amount of training time and data to reach reasonable\nperformance, making it inapplicable to real-world settings where data is\nexpensive. In this work, we improve data efficiency in deep RL by addressing\none of the two learning goals, feature learning. We leverage supervised\nlearning to pre-train on a small set of non-expert human demonstrations and\nempirically evaluate our approach using the asynchronous advantage actor-critic\nalgorithms (A3C) in the Atari domain. Our results show significant improvements\nin learning speed, even when the provided demonstration is noisy and of low\nquality.","url_abs":"http://arxiv.org/abs/1812.08904v1","url_pdf":"http://arxiv.org/pdf/1812.08904v1.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":"pre-training-with-non-expert-human","repo_url":"https://github.com/gabrieledcjr/DeepRL","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"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":{"atlas_url":"https://app.syntology.ai/?focus=1812.08904","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}