Timezone: »
Poster
Agnostic Reinforcement Learning with Low-Rank MDPs and Rich Observations
Ayush Sekhari · Christoph Dann · Mehryar Mohri · Yishay Mansour · Karthik Sridharan
There have been many recent advances on provably efficient Reinforcement Learning (RL) in problems with rich observation spaces. However, all these works share a strong realizability assumption about the optimal value function of the true MDP. Such realizability assumptions are often too strong to hold in practice. In this work, we consider the more realistic setting of agnostic RL with rich observation spaces and a fixed class of policies $\Pi$ that may not contain any near-optimal policy. We provide an algorithm for this setting whose error is bounded in terms of the rank $d$ of the underlying MDP. Specifically, our algorithm enjoys a sample complexity bound of $\widetilde{O}\left((H^{4d} K^{3d} \log |\Pi|)/\epsilon^2\right)$ where $H$ is the length of episodes, $K$ is the number of actions and $\epsilon>0$ is the desired sub-optimality. We also provide a nearly matching lower bound for this agnostic setting that shows that the exponential dependence on rank is unavoidable, without further assumptions.
Author Information
Ayush Sekhari (Cornell University)
Christoph Dann (Google Research)
Mehryar Mohri (Google Research & Courant Institute of Mathematical Sciences)
Yishay Mansour (Tel Aviv University & Google)
Karthik Sridharan (Cornell)
Related Events (a corresponding poster, oral, or spotlight)
-
2021 Spotlight: Agnostic Reinforcement Learning with Low-Rank MDPs and Rich Observations »
Dates n/a. Room
More from the Same Authors
-
2021 Spotlight: On the Existence of The Adversarial Bayes Classifier »
Pranjal Awasthi · Natalie Frank · Mehryar Mohri -
2021 Spotlight: Beyond Value-Function Gaps: Improved Instance-Dependent Regret Bounds for Episodic Reinforcement Learning »
Christoph Dann · Teodor Vanislavov Marinov · Mehryar Mohri · Julian Zimmert -
2021 Spotlight: Calibration and Consistency of Adversarial Surrogate Losses »
Pranjal Awasthi · Natalie Frank · Anqi Mao · Mehryar Mohri · Yutao Zhong -
2022 : A Theory of Learning with Competing Objectives and User Feedback »
Pranjal Awasthi · Corinna Cortes · Yishay Mansour · Mehryar Mohri -
2022 : AdaME: Adaptive learning of multisource adaptationensembles »
Scott Yak · Javier Gonzalvo · Mehryar Mohri · Corinna Cortes -
2022 : A Theory of Learning with Competing Objectives and User Feedback »
Pranjal Awasthi · Corinna Cortes · Yishay Mansour · Mehryar Mohri -
2022 : Finding Safe Zones of Markov Decision Processes Policies »
Michal Moshkovitz · Lee Cohen · Yishay Mansour -
2022 Spotlight: Lightning Talks 6A-2 »
Yichuan Mo · Botao Yu · Gang Li · Zezhong Xu · Haoran Wei · Arsene Fansi Tchango · Raef Bassily · Haoyu Lu · Qi Zhang · Songming Liu · Mingyu Ding · Peiling Lu · Yifei Wang · Xiang Li · Dongxian Wu · Ping Guo · Wen Zhang · Hao Zhongkai · Mehryar Mohri · Rishab Goel · Yisen Wang · Yifei Wang · Yangguang Zhu · Zhi Wen · Ananda Theertha Suresh · Chengyang Ying · Yujie Wang · Peng Ye · Rui Wang · Nanyi Fei · Hui Chen · Yiwen Guo · Wei Hu · Chenglong Liu · Julien Martel · Yuqi Huo · Wu Yichao · Hang Su · Yisen Wang · Peng Wang · Huajun Chen · Xu Tan · Jun Zhu · Ding Liang · Zhiwu Lu · Joumana Ghosn · Shanshan Zhang · Wei Ye · Ze Cheng · Shikun Zhang · Tao Qin · Tie-Yan Liu -
2022 Spotlight: Differentially Private Learning with Margin Guarantees »
Raef Bassily · Mehryar Mohri · Ananda Theertha Suresh -
2022 : A Theory of Learning with Competing Objectives and User Feedback »
Pranjal Awasthi · Corinna Cortes · Yishay Mansour · Mehryar Mohri -
2022 : Invited Talk #1, Differentially Private Learning with Margin Guarantees, Mehryar Mohri »
Mehryar Mohri -
2022 Poster: Benign Underfitting of Stochastic Gradient Descent »
Tomer Koren · Roi Livni · Yishay Mansour · Uri Sherman -
2022 Poster: A Characterization of Semi-Supervised Adversarially Robust PAC Learnability »
Idan Attias · Steve Hanneke · Yishay Mansour -
2022 Poster: Near-Optimal Regret for Adversarial MDP with Delayed Bandit Feedback »
Tiancheng Jin · Tal Lancewicki · Haipeng Luo · Yishay Mansour · Aviv Rosenberg -
2022 Poster: Best of Both Worlds Model Selection »
Aldo Pacchiano · Christoph Dann · Claudio Gentile -
2022 Poster: Multi-Class $H$-Consistency Bounds »
Pranjal Awasthi · Anqi Mao · Mehryar Mohri · Yutao Zhong -
2022 Poster: Stochastic Online Learning with Feedback Graphs: Finite-Time and Asymptotic Optimality »
Teodor Vanislavov Marinov · Mehryar Mohri · Julian Zimmert -
2022 Poster: Fair Wrapping for Black-box Predictions »
Alexander Soen · Ibrahim Alabdulmohsin · Sanmi Koyejo · Yishay Mansour · Nyalleng Moorosi · Richard Nock · Ke Sun · Lexing Xie -
2022 Poster: Differentially Private Learning with Margin Guarantees »
Raef Bassily · Mehryar Mohri · Ananda Theertha Suresh -
2021 Poster: SGD: The Role of Implicit Regularization, Batch-size and Multiple-epochs »
Ayush Sekhari · Karthik Sridharan · Satyen Kale -
2021 Poster: A Provably Efficient Model-Free Posterior Sampling Method for Episodic Reinforcement Learning »
Christoph Dann · Mehryar Mohri · Tong Zhang · Julian Zimmert -
2021 Poster: On the Existence of The Adversarial Bayes Classifier »
Pranjal Awasthi · Natalie Frank · Mehryar Mohri -
2021 Poster: Beyond Value-Function Gaps: Improved Instance-Dependent Regret Bounds for Episodic Reinforcement Learning »
Christoph Dann · Teodor Vanislavov Marinov · Mehryar Mohri · Julian Zimmert -
2021 Poster: Minimax Regret for Stochastic Shortest Path »
Alon Cohen · Yonathan Efroni · Yishay Mansour · Aviv Rosenberg -
2021 Oral: Optimal Rates for Random Order Online Optimization »
Uri Sherman · Tomer Koren · Yishay Mansour -
2021 Poster: Learning with User-Level Privacy »
Daniel Levy · Ziteng Sun · Kareem Amin · Satyen Kale · Alex Kulesza · Mehryar Mohri · Ananda Theertha Suresh -
2021 Poster: Boosting with Multiple Sources »
Corinna Cortes · Mehryar Mohri · Dmitry Storcheus · Ananda Theertha Suresh -
2021 Poster: Breaking the centralized barrier for cross-device federated learning »
Sai Praneeth Karimireddy · Martin Jaggi · Satyen Kale · Mehryar Mohri · Sashank Reddi · Sebastian Stich · Ananda Theertha Suresh -
2021 Poster: Optimal Rates for Random Order Online Optimization »
Uri Sherman · Tomer Koren · Yishay Mansour -
2021 Poster: Oracle-Efficient Regret Minimization in Factored MDPs with Unknown Structure »
Aviv Rosenberg · Yishay Mansour -
2021 Poster: Differentially Private Multi-Armed Bandits in the Shuffle Model »
Jay Tenenbaum · Haim Kaplan · Yishay Mansour · Uri Stemmer -
2021 Poster: ROI Maximization in Stochastic Online Decision-Making »
Nicolò Cesa-Bianchi · Tom Cesari · Yishay Mansour · Vianney Perchet -
2021 Poster: Neural Active Learning with Performance Guarantees »
Zhilei Wang · Pranjal Awasthi · Christoph Dann · Ayush Sekhari · Claudio Gentile -
2021 Poster: Remember What You Want to Forget: Algorithms for Machine Unlearning »
Ayush Sekhari · Jayadev Acharya · Gautam Kamath · Ananda Theertha Suresh -
2021 Poster: Dueling Bandits with Team Comparisons »
Lee Cohen · Ulrike Schmidt-Kraepelin · Yishay Mansour -
2021 Poster: Calibration and Consistency of Adversarial Surrogate Losses »
Pranjal Awasthi · Natalie Frank · Anqi Mao · Mehryar Mohri · Yutao Zhong -
2020 Poster: Adapting to Misspecification in Contextual Bandits »
Dylan Foster · Claudio Gentile · Mehryar Mohri · Julian Zimmert -
2020 Poster: Agnostic Learning with Multiple Objectives »
Corinna Cortes · Mehryar Mohri · Javier Gonzalvo · Dmitry Storcheus -
2020 Poster: Reinforcement Learning with Feedback Graphs »
Christoph Dann · Yishay Mansour · Mehryar Mohri · Ayush Sekhari · Karthik Sridharan -
2020 Poster: PAC-Bayes Learning Bounds for Sample-Dependent Priors »
Pranjal Awasthi · Satyen Kale · Stefani Karp · Mehryar Mohri -
2019 : Mehryar Mohri, "Learning with Sample-Dependent Hypothesis Sets" »
Mehryar Mohri -
2019 Poster: Learning GANs and Ensembles Using Discrepancy »
Ben Adlam · Corinna Cortes · Mehryar Mohri · Ningshan Zhang -
2019 Poster: Bandits with Feedback Graphs and Switching Costs »
Raman Arora · Teodor Vanislavov Marinov · Mehryar Mohri -
2019 Poster: Regularized Gradient Boosting »
Corinna Cortes · Mehryar Mohri · Dmitry Storcheus -
2019 Poster: Hypothesis Set Stability and Generalization »
Dylan Foster · Spencer Greenberg · Satyen Kale · Haipeng Luo · Mehryar Mohri · Karthik Sridharan -
2018 Poster: Policy Regret in Repeated Games »
Raman Arora · Michael Dinitz · Teodor Vanislavov Marinov · Mehryar Mohri -
2018 Poster: Efficient Gradient Computation for Structured Output Learning with Rational and Tropical Losses »
Corinna Cortes · Vitaly Kuznetsov · Mehryar Mohri · Dmitry Storcheus · Scott Yang -
2018 Poster: On Oracle-Efficient PAC RL with Rich Observations »
Christoph Dann · Nan Jiang · Akshay Krishnamurthy · Alekh Agarwal · John Langford · Robert Schapire -
2018 Poster: Algorithms and Theory for Multiple-Source Adaptation »
Judy Hoffman · Mehryar Mohri · Ningshan Zhang -
2018 Poster: Uniform Convergence of Gradients for Non-Convex Learning and Optimization »
Dylan Foster · Ayush Sekhari · Karthik Sridharan -
2018 Spotlight: On Oracle-Efficient PAC RL with Rich Observations »
Christoph Dann · Nan Jiang · Akshay Krishnamurthy · Alekh Agarwal · John Langford · Robert Schapire -
2017 : Mehryar Mohri (NYU) on Tight Learning Bounds for Multi-Class Classification »
Mehryar Mohri -
2017 : (Invited Talk) Mehryar Mohri: Regret minimization against strategic buyers. »
Mehryar Mohri -
2017 Workshop: Learning in the Presence of Strategic Behavior »
Nika Haghtalab · Yishay Mansour · Tim Roughgarden · Vasilis Syrgkanis · Jennifer Wortman Vaughan -
2017 Poster: Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues »
Noga Alon · Moshe Babaioff · Yannai A. Gonczarowski · Yishay Mansour · Shay Moran · Amir Yehudayoff -
2017 Poster: Discriminative State Space Models »
Vitaly Kuznetsov · Mehryar Mohri -
2017 Poster: Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning »
Christoph Dann · Tor Lattimore · Emma Brunskill -
2017 Spotlight: Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning »
Christoph Dann · Tor Lattimore · Emma Brunskill -
2017 Spotlight: Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues »
Noga Alon · Moshe Babaioff · Yannai A. Gonczarowski · Yishay Mansour · Shay Moran · Amir Yehudayoff -
2017 Poster: Multi-Armed Bandits with Metric Movement Costs »
Tomer Koren · Roi Livni · Yishay Mansour -
2017 Poster: Online Learning with Transductive Regret »
Scott Yang · Mehryar Mohri -
2017 Poster: Parameter-Free Online Learning via Model Selection »
Dylan J Foster · Satyen Kale · Mehryar Mohri · Karthik Sridharan -
2017 Spotlight: Parameter-Free Online Learning via Model Selection »
Dylan J Foster · Satyen Kale · Mehryar Mohri · Karthik Sridharan -
2017 Spotlight: Online Learning with Transductive Regret »
Scott Yang · Mehryar Mohri -
2016 Poster: (Withdrawn)Only H is left: Near-tight Episodic PAC RL »
Christoph Dann · Emma Brunskill -
2016 Poster: Structured Prediction Theory Based on Factor Graph Complexity »
Corinna Cortes · Vitaly Kuznetsov · Mehryar Mohri · Scott Yang -
2016 Poster: Boosting with Abstention »
Corinna Cortes · Giulia DeSalvo · Mehryar Mohri -
2016 Poster: Optimistic Bandit Convex Optimization »
Scott Yang · Mehryar Mohri -
2016 Tutorial: Theory and Algorithms for Forecasting Non-Stationary Time Series »
Vitaly Kuznetsov · Mehryar Mohri -
2015 : A Theory of Multiple Source Adaptation »
Mehryar Mohri -
2015 : Learning Theory and Algorithms for Time Series »
Mehryar Mohri -
2015 Poster: Revenue Optimization against Strategic Buyers »
Mehryar Mohri · Andres Munoz -
2015 Poster: Learning Theory and Algorithms for Forecasting Non-stationary Time Series »
Vitaly Kuznetsov · Mehryar Mohri -
2015 Oral: Learning Theory and Algorithms for Forecasting Non-stationary Time Series »
Vitaly Kuznetsov · Mehryar Mohri -
2015 Poster: Sample Complexity of Episodic Fixed-Horizon Reinforcement Learning »
Christoph Dann · Emma Brunskill -
2015 Poster: The Human Kernel »
Andrew Wilson · Christoph Dann · Chris Lucas · Eric Xing -
2015 Spotlight: The Human Kernel »
Andrew Wilson · Christoph Dann · Chris Lucas · Eric Xing -
2014 Workshop: Second Workshop on Transfer and Multi-Task Learning: Theory meets Practice »
Urun Dogan · Tatiana Tommasi · Yoshua Bengio · Francesco Orabona · Marius Kloft · Andres Munoz · Gunnar Rätsch · Hal Daumé III · Mehryar Mohri · Xuezhi Wang · Daniel Hernández-lobato · Song Liu · Thomas Unterthiner · Pascal Germain · Vinay P Namboodiri · Michael Goetz · Christopher Berlind · Sigurd Spieckermann · Marta Soare · Yujia Li · Vitaly Kuznetsov · Wenzhao Lian · Daniele Calandriello · Emilie Morvant -
2014 Workshop: NIPS Workshop on Transactional Machine Learning and E-Commerce »
David Parkes · David H Wolpert · Jennifer Wortman Vaughan · Jacob D Abernethy · Amos Storkey · Mark Reid · Ping Jin · Nihar Bhadresh Shah · Mehryar Mohri · Luis E Ortiz · Robin Hanson · Aaron Roth · Satyen Kale · Sebastien Lahaie -
2014 Poster: Optimal Regret Minimization in Posted-Price Auctions with Strategic Buyers »
Mehryar Mohri · Andres Munoz -
2014 Poster: Multi-Class Deep Boosting »
Vitaly Kuznetsov · Mehryar Mohri · Umar Syed -
2014 Spotlight: Optimal Regret Minimization in Posted-Price Auctions with Strategic Buyers »
Mehryar Mohri · Andres Munoz -
2014 Session: Oral Session 6 »
Mehryar Mohri -
2014 Poster: Conditional Swap Regret and Conditional Correlated Equilibrium »
Mehryar Mohri · Scott Yang -
2013 Poster: Learning Kernels Using Local Rademacher Complexity »
Corinna Cortes · Marius Kloft · Mehryar Mohri -
2013 Spotlight: Learning Kernels Using Local Rademacher Complexity »
Corinna Cortes · Marius Kloft · Mehryar Mohri -
2012 Poster: Accuracy at the Top »
Stephen Boyd · Corinna Cortes · Mehryar Mohri · Ana Radovanovic -
2012 Poster: Spectral Learning of General Weighted Automata via Constrained Matrix Completion »
Borja Balle · Mehryar Mohri -
2012 Oral: Spectral Learning of General Weighted Automata via Constrained Matrix Completion »
Borja Balle · Mehryar Mohri -
2011 Workshop: Sparse Representation and Low-rank Approximation »
Ameet S Talwalkar · Lester W Mackey · Mehryar Mohri · Michael W Mahoney · Francis Bach · Mike Davies · Remi Gribonval · Guillaume R Obozinski -
2010 Workshop: Low-rank Methods for Large-scale Machine Learning »
Arthur Gretton · Michael W Mahoney · Mehryar Mohri · Ameet S Talwalkar -
2010 Poster: Learning Bounds for Importance Weighting »
Corinna Cortes · Yishay Mansour · Mehryar Mohri -
2009 Poster: Efficient Large-Scale Distributed Training of Conditional Maximum Entropy Models »
Gideon S Mann · Ryan McDonald · Mehryar Mohri · Nathan Silberman · Dan Walker -
2009 Poster: Ensemble Nystrom Method »
Sanjiv Kumar · Mehryar Mohri · Ameet S Talwalkar -
2009 Spotlight: Efficient Large-Scale Distributed Training of Conditional Maximum Entropy Models »
Gideon S Mann · Ryan McDonald · Mehryar Mohri · Nathan Silberman · Dan Walker -
2009 Poster: Learning Non-Linear Combinations of Kernels »
Corinna Cortes · Mehryar Mohri · Afshin Rostamizadeh -
2009 Poster: Polynomial Semantic Indexing »
Bing Bai · Jason E Weston · David Grangier · Ronan Collobert · Kunihiko Sadamasa · Yanjun Qi · Corinna Cortes · Mehryar Mohri -
2008 Workshop: Kernel Learning: Automatic Selection of Optimal Kernels »
Corinna Cortes · Arthur Gretton · Gert Lanckriet · Mehryar Mohri · Afshin Rostamizadeh -
2008 Poster: Domain Adaptation with Multiple Sources »
Yishay Mansour · Mehryar Mohri · Afshin Rostamizadeh -
2008 Spotlight: Domain Adaptation with Multiple Sources »
Yishay Mansour · Mehryar Mohri · Afshin Rostamizadeh -
2008 Poster: Rademacher Complexity Bounds for Non-I.I.D. Processes »
Mehryar Mohri · Afshin Rostamizadeh -
2007 Poster: Stability Bounds for Non-i.i.d. Processes »
Mehryar Mohri · Afshin Rostamizadeh -
2006 Poster: On Transductive Regression »
Corinna Cortes · Mehryar Mohri