{"id":"arxiv:1910.01708","title":"Benchmarking Batch Deep Reinforcement Learning Algorithms","date":"2019-10-03","section":"Reinforcement learning","quality":-0.5589,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/","source_url":"https://arxiv.org/abs/1910.01708","text_available":true,"authors":["Scott Fujimoto","Edoardo Conti","Mohammad Ghavamzadeh","Joelle Pineau"],"outgoing":[{"source":"arxiv:1910.01708","target":"arxiv:1509.06461","evidence":"title","id":"arxiv:1509.06461","title":"Deep Reinforcement Learning with Double Q-learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a2f115ac879cd2efb9548baf3ecdcb925d2492f03b4b0940720f6e5b0840715b/"},{"source":"arxiv:1910.01708","target":"arxiv:1707.06887","evidence":"title","id":"arxiv:1707.06887","title":"A Distributional Perspective on Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/96b040bb771d28ca3254b5519a2191c3e958a35dfdd60075344ac86af6a590e1/"},{"source":"arxiv:1910.01708","target":"arxiv:1710.02298","evidence":"id","id":"arxiv:1710.02298","title":"Rainbow: Combining Improvements in Deep Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4183a5aecdb6558733bd0b2028e6197904c8b46173181bf6c8270fb6cbcfc973/"},{"source":"arxiv:1910.01708","target":"arxiv:1802.09477","evidence":"title","id":"arxiv:1802.09477","title":"Addressing Function Approximation Error in Actor-Critic Methods","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/5d1a410c86d85da493735fae67eb1b2689350e5ac449d39717deafaf3904e7c1/"}],"incoming":[{"source":"arxiv:2106.06860","target":"arxiv:1910.01708","evidence":"id","id":"arxiv:2106.06860","title":"A Minimalist Approach to Offline Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/00a35c7a58d8775f23035efbd1cb0a5f540440d47b8a996be8a4fa0da2da3e0e/"}],"text_parts":[{"start":0,"end":15959,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html"},{"start":15959,"end":31411,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html"},{"start":31411,"end":35865,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-3.html"}],"contents":[{"title":"Benchmarking Batch Deep Reinforcement Learning Algorithms","offset":0,"level":1,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html#text"},{"title":"Abstract","offset":61,"level":6,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html#text"},{"title":"1 Introduction","offset":960,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html#text"},{"title":"2 Preliminaries","offset":5748,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html#text"},{"title":"3 Batch Deep Reinforcement Learning Algorithms","offset":11152,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-1.html#text"},{"title":"4 Discrete Batch-Constrained Deep Q-learning","offset":21991,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html#text"},{"title":"5 Experiments","offset":24980,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html#text"},{"title":"6 Conclusion","offset":30242,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html#text"},{"title":"Appendix A Experimental Details","offset":31129,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html#text"},{"title":"A.1 Atari Preprocessing.","offset":31165,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-2.html#text"},{"title":"A.2 Architecture and Hyper-parameters.","offset":32789,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-3.html#text"},{"title":"Instructions for reporting errors","offset":34901,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/part-3.html#text"}],"text_code_points":35865,"source_markdown":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/ce9cb9b764eb7fdaf1aa766cb24e6abaec20a00445ad5af5a6dacb5af3ad3ddc/source.md"}