Regularization and Variance-Weighted Regression Achieves Minimax Optimality in Linear MDPs: Theory and Practice

Toshinori Kitamura, Tadashi Kozuno, Yunhao Tang, Nino Vieillard, Michal Valko, Wenhao Yang, Jincheng Mei, Pierre Ménard, Mohammad Gheshlaghi Azar, Rémi Munos, Olivier Pietquin, Matthieu Geist, Csaba Szepesvári, Wataru Kumagai, Yutaka Matsuo. Regularization and Variance-Weighted Regression Achieves Minimax Optimality in Linear MDPs: Theory and Practice. In Andreas Krause 0001, Emma Brunskill, KyungHyun Cho, Barbara Engelhardt, Sivan Sabato, Jonathan Scarlett, editors, International Conference on Machine Learning, ICML 2023, 23-29 July 2023, Honolulu, Hawaii, USA. Volume 202 of Proceedings of Machine Learning Research, pages 17135-17175, PMLR, 2023. [doi]

@inproceedings{KitamuraKTVVYMM23,
  title = {Regularization and Variance-Weighted Regression Achieves Minimax Optimality in Linear MDPs: Theory and Practice},
  author = {Toshinori Kitamura and Tadashi Kozuno and Yunhao Tang and Nino Vieillard and Michal Valko and Wenhao Yang and Jincheng Mei and Pierre Ménard and Mohammad Gheshlaghi Azar and Rémi Munos and Olivier Pietquin and Matthieu Geist and Csaba Szepesvári and Wataru Kumagai and Yutaka Matsuo},
  year = {2023},
  url = {https://proceedings.mlr.press/v202/kitamura23a.html},
  researchr = {https://researchr.org/publication/KitamuraKTVVYMM23},
  cites = {0},
  citedby = {0},
  pages = {17135-17175},
  booktitle = {International Conference on Machine Learning, ICML 2023, 23-29 July 2023, Honolulu, Hawaii, USA},
  editor = {Andreas Krause 0001 and Emma Brunskill and KyungHyun Cho and Barbara Engelhardt and Sivan Sabato and Jonathan Scarlett},
  volume = {202},
  series = {Proceedings of Machine Learning Research},
  publisher = {PMLR},
}