Papers
Topics
Authors
Recent
Detailed Answer
Quick Answer
Concise responses based on abstracts only
Detailed Answer
Well-researched responses based on abstracts and relevant paper content.
Custom Instructions Pro
Preferences or requirements that you'd like Emergent Mind to consider when generating responses
Gemini 2.5 Flash
Gemini 2.5 Flash 91 tok/s
Gemini 2.5 Pro 56 tok/s Pro
GPT-5 Medium 29 tok/s Pro
GPT-5 High 29 tok/s Pro
GPT-4o 108 tok/s Pro
Kimi K2 214 tok/s Pro
GPT OSS 120B 470 tok/s Pro
Claude Sonnet 4 36 tok/s Pro
2000 character limit reached

Grid-Mapping Pseudo-Count Constraint for Offline Reinforcement Learning (2404.02545v2)

Published 3 Apr 2024 in cs.LG and cs.AI

Abstract: Offline reinforcement learning learns from a static dataset without interacting with environments, which ensures security and thus owns a good application prospect. However, directly applying naive reinforcement learning algorithm usually fails in an offline environment due to inaccurate Q value approximation caused by out-of-distribution (OOD) state-actions. It is an effective way to solve this problem by penalizing the Q-value of OOD state-actions. Among the methods of punishing OOD state-actions, count-based methods have achieved good results in discrete domains in a simple form. Inspired by it, a novel pseudo-count method for continuous domains called Grid-Mapping Pseudo-Count method (GPC) is proposed by extending the count-based method from discrete to continuous domains. Firstly, the continuous state and action space are mapped to discrete space using Grid-Mapping, then the Q-values of OOD state-actions are constrained through pseudo-count. Secondly, the theoretical proof is given to show that GPC can obtain appropriate uncertainty constraints under fewer assumptions than other pseudo-count methods. Thirdly, GPC is combined with Soft Actor-Critic algorithm (SAC) to get a new algorithm called GPC-SAC. Lastly, experiments on D4RL datasets are given to show that GPC-SAC has better performance and less computational cost than other algorithms that constrain the Q-value.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (41)
  1. OpenAI, “Gpt-4 technical report,” 2023.
  2. O. Vinyals, I. Babuschkin, W. M. Czarnecki, M. Mathieu, A. Dudzik, J. Chung, D. H. Choi, R. Powell, T. Ewalds, P. Georgiev, J. Oh, D. Horgan, M. Kroiss, I. Danihelka, A. Huang, L. Sifre, T. Cai, J. P. Agapiou, M. Jaderberg, A. S. Vezhnevets, R. Leblond, T. Pohlen, V. Dalibard, D. Budden, Y. Sulsky, J. Molloy, T. L. Paine, C. Gulcehre, Z. Wang, T. Pfaff, Y. Wu, R. Ring, D. Yogatama, D. Wünsch, K. McKinney, O. Smith, T. Schaul, T. Lillicrap, K. Kavukcuoglu, D. Hassabis, C. Apps, and D. Silver, “Grandmaster level in starcraft ii using multi-agent reinforcement learning,” Nature, vol. 575, no. 7782, pp. 350–354, 2019.
  3. E. Kaufmann, L. Bauersfeld, A. Loquercio, M. Mx¨¨𝑥\ddot{x}over¨ start_ARG italic_x end_ARGller, V. Koltun, and D. Scaramuzza, “Champion-level drone racing using deep reinforcement learning,” Nature, vol. 620, no. 7976, pp. 982–987, 2023.
  4. C. Yu, J. Liu, and S. Nemati, “Reinforcement learning in healthcare: A survey,” ACM Computing Surveys (CSUR), vol. 55, no. 1, pp. 1–36, 2021.
  5. S. Levine, A. Kumar, G. Tucker, and J. Fu, “Offline reinforcement learning: Tutorial, review, and perspectives on open problems,” arXiv preprint arXiv:2005.01643, 2020.
  6. S. Fujimoto, D. Meger, and D. Precup, “Off-policy deep reinforcement learning without exploration,” in International conference on machine learning.   PMLR, 2019, pp. 2052–2062.
  7. R. Agarwal, D. Schuurmans, and M. Norouzi, “An optimistic perspective on offline reinforcement learning,” in International Conference on Machine Learning.   PMLR, 2020, pp. 104–114.
  8. R. F. Prudencio, M. R. O. A. Maximo, and E. L. Colombini, “A survey on offline reinforcement learning: Taxonomy, review, and open problems,” IEEE Transactions on Neural Networks and Learning Systems, 2023.
  9. A. Kumar, J. Fu, M. Soh, G. Tucker, and S. Levine, “Stabilizing off-policy q-learning via bootstrapping error reduction,” Advances in Neural Information Processing Systems, vol. 32, 2019.
  10. A. Kumar, A. Zhou, G. Tucker, and S. Levine, “Conservative q-learning for offline reinforcement learning,” Advances in Neural Information Processing Systems, vol. 33, pp. 1179–1191, 2020.
  11. C. Bai, L. Wang, Z. Yang, Z. Deng, A. Garg, P. Liu, and Z. Wang, “Pessimistic bootstrapping for uncertainty-driven offline reinforcement learning,” arXiv preprint arXiv:2202.11566, 2022.
  12. C. Gulcehre, Z. Wang, A. Novikov, T. L. Paine, S. G. Colmenarejo, K. Zolna, R. Agarwal, J. Merel, D. Mankowitz, C. Paduraru, G. Dulac-Arnold, J. Li, M. Norouzi, M. Hoffman, O. Nachum, G. Tucker, N. Heess, and N. de Freitas, “Rl unplugged: A suite of benchmarks for offline reinforcement learning,” Advances in Neural Information Processing Systems, vol. 33, pp. 7248–7259, 2020.
  13. I. Osband, J. Aslanides, and A. Cassirer, “Randomized prior functions for deep reinforcement learning,” Advances in Neural Information Processing Systems, vol. 31, 2018.
  14. S. Rezaeifar, R. Dadashi, N. Vieillard, L. Hussenot, O. Bachem, O. Pietquin, and M. Geist, “Offline reinforcement learning as anti-exploration,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 36, no. 7, 2022, pp. 8106–8114.
  15. J. Hong, A. Kumar, and S. Levine, “Confidence-conditioned value functions for offline reinforcement learning,” arXiv preprint arXiv:2212.04607, 2022.
  16. J. Fu, A. Kumar, O. Nachum, G. Tucker, and S. Levine, “D4rl: Datasets for deep data-driven reinforcement learning,” arXiv preprint arXiv:2004.07219, 2020.
  17. T. Haarnoja, A. Zhou, P. Abbeel, and S. Levine, “Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor,” in International conference on machine learning.   PMLR, 2018, pp. 1861–1870.
  18. V. Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, S. Petersen, C. Beattie, A. Sadik, I. Antonoglou, H. King, D. Kumaran, D. Wierstra, S. Legg, and D. Hassabis, “Human-level control through deep reinforcement learning,” nature, vol. 518, no. 7540, pp. 529–533, 2015.
  19. C.-J. Hoel, T. Tram, and J. Sjöberg, “Reinforcement learning with uncertainty estimation for tactical decision-making in intersections,” in 2020 IEEE 23rd international conference on intelligent transportation systems (ITSC).   IEEE, 2020, pp. 1–7.
  20. W. R. Clements, B. V. Delft, B.-M. Robaglia, R. B. Slaoui, and S. Toth, “Estimating risk and uncertainty in deep reinforcement learning,” arXiv preprint arXiv:1905.09638, 2019.
  21. M. Janner, J. Fu, M. Zhang, and S. Levine, “When to trust your model: Model-based policy optimization,” Advances in neural information processing systems, vol. 32, 2019.
  22. T. Yu, G. Thomas, L. Yu, S. Ermon, J. Zou, S. Levine, C. Finn, and T. Ma, “Mopo: Model-based offline policy optimization,” Advances in Neural Information Processing Systems, vol. 33, pp. 14 129–14 142, 2020.
  23. T. Yu, A. Kumar, R. Rafailov, A. Rajeswaran, S. Levine, and C. Fin, “Combo: Conservative offline model-based policy optimization,” Advances in neural information processing systems, vol. 34, pp. 28 954–28 967, 2021.
  24. V. Mai, K. Mani, and L. Paull, “Sample efficient deep reinforcement learning via uncertainty estimation,” arXiv preprint arXiv:2201.01666, 2022.
  25. M. G. Bellemare, S. Srinivasan, G. Ostrovski, T. Schaul, D. Saxton, and R. Munos, “Unifying count-based exploration and intrinsic motivation,” Advances in neural information processing systems, vol. 29, 2016.
  26. A. Agarwal, S. Kakade, A. Krishnamurthy, and W. Sun, “Flambe: Structural complexity and representation learning of low rank mdps,” Advances in neural information processing systems, vol. 33, pp. 20 095–20 107, 2020.
  27. R. T. McAllister and C. E. Rasmussen, “Improving pilco with bayesian neural network dynamics models,” in Data-efficient machine learning workshop, ICML, vol. 4, no. 34, 2016, p. 25.
  28. Y. J. Ma, A. Shen, O. Bastani, and D. Jayaraman, “Conservative and adaptive penalty for model-based safe reinforcement learning,” in Proceedings of the AAAI conference on artificial intelligence, vol. 36, no. 5, 2022, pp. 5404–5412.
  29. Z. Zhu, E. Bıyık, and D. Sadigh, “Multi-agent safe planning with gaussian processes,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).   IEEE, 2020, pp. 6260–6267.
  30. R. Kidambi, A. Rajeswaran, P. Netrapalli, and T. Joachims, “Morel: Model-based offline reinforcement learning,” Advances in neural information processing systems, vol. 33, pp. 21 810–21 823, 2020.
  31. Y. Wu, S. Zhai, N. Srivastava, J. Susskind, J. Zhang, R. Salakhutdinov, and H. Goh, “Uncertainty weighted actor-critic for offline reinforcement learning,” arXiv preprint arXiv:2105.08140, 2021.
  32. G. An, S. Moon, J.-H. Kim, and H. O. Song, “Uncertainty-based offline reinforcement learning with diversified q-ensemble,” Advances in neural information processing systems, vol. 34, pp. 7436–7447, 2021.
  33. G. Ostrovski, M. G. Bellemare, A. Oord, and R. Munos, “Count-based exploration with neural density models,” in International conference on machine learning.   PMLR, 2017, pp. 2721–2730.
  34. H. Tang, R. Houthooft, D. Foote, A. Stooke, X. Chen, Y. Duan, J. Schulman, F. D. Turck, and P. Abbeel, “# exploration: A study of count-based exploration for deep reinforcement learning,” Advances in neural information processing systems, vol. 30, 2017.
  35. J. Choi, Y. Guo, M. Moczulski, J. Oh, N. Wu, M. Norouzi, and H. Lee, “Contingency-aware exploration in reinforcement learning,” arXiv preprint arXiv:1811.01483, 2018.
  36. W. Hoeffding, “Probability inequalities for sums of bounded random variables,” The collected works of Wassily Hoeffding, pp. 409–426, 1994.
  37. C. Jin, Z. Yang, Z. Wang, and M. I. Jordan, “Provably efficient reinforcement learning with linear function approximation,” in Conference on Learning Theory.   PMLR, 2020, pp. 2137–2143.
  38. R. Wang, S. S. Du, L. Yang, and R. R. Salakhutdinov, “On reward-free reinforcement learning with linear function approximation,” Advances in neural information processing systems, vol. 33, pp. 17 816–17 826, 2020.
  39. Y. Abbasi-Yadkori, D. Pál, and C. Szepesvári, “Improved algorithms for linear stochastic bandits,” Advances in neural information processing systems, vol. 24, 2011.
  40. I. Kostrikov, A. Nair, and S. Levine, “Offline reinforcement learning with implicit q-learning,” ArXiv, vol. abs/2110.06169, 2021.
  41. Y. Gal, R. McAllister, and C. E. Rasmussen, “Improving pilco with bayesian neural network dynamics models,” in Data-efficient machine learning workshop, ICML, vol. 4, no. 34, 2016, p. 25.
List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
Dice Question Streamline Icon: https://streamlinehq.com

Follow-Up Questions

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Authors (2)

  1. Yi Shen
  2. Hanyan Huang
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets