Distributionally Robust

Distributionally robust methods aim to create machine learning models and control systems that perform well even when the data or environment differs from the training conditions. Current research focuses on developing algorithms and theoretical frameworks for various applications, including reinforcement learning, inverse reinforcement learning, and classification, often employing techniques like Wasserstein distance, KL divergence, and total variation distance to define uncertainty sets. This work is significant because it addresses the critical issue of robustness in machine learning and control, leading to more reliable and dependable systems in real-world scenarios such as autonomous driving and robotics. The development of efficient algorithms with provable guarantees is a major focus of ongoing research.

Papers

October 31, 2024

PSL: Rethinking and Improving Softmax Loss from Pairwise Perspective for Recommendation
Weiqin Yang, Jiawei Chen, Xin Xin, Sheng Zhou, Binbin Hu, Yan Feng, Chun Chen, Can Wang
Loss Function Recommender System Effective Recommendation Distributionally Robust Softmax Loss Pairwise Approach

October 17, 2024

Gradual Domain Adaptation via Manifold-Constrained Distributionally Robust Optimization
Amir Hossein Saberi, Amir Najafi, Ala Emrani, Amin Behjati, Yasaman Zolfimoselo, Mahdi Shadrooy, Abolfazl Motahari, Babak H. Khalaj
Domain Shift Robust Optimization Distributionally Robust Gradual Domain Adaptation

September 30, 2024

Upper and Lower Bounds for Distributionally Robust Off-Dynamics Reinforcement Learning
Zhishuai Liu, Weixin Wang, Pan Xu
Lower Bound Distributionally Robust Dual Optimization Policy Switching

September 22, 2024

Distributionally Robust Inverse Reinforcement Learning for Identifying Multi-Agent Coordinated Sensing
Luke Snow, Vikram Krishnamurthy
Inverse Reinforcement Learning Robust Estimation Multi Agent Coordination Distributionally Robust

July 26, 2024

Addressing Behavior Model Inaccuracies for Safe Motion Control in Uncertain Dynamic Environments
Minjun Sung, Hunmin Kim, Naira Hovakimyan
State Estimation Distributionally Robust Behavior Model Autonomous Driving Simulation Safe Motion Uncertain Dynamic Environment

July 23, 2024

Distribution-Aware Robust Learning from Long-Tailed Data with Noisy Labels
Jae Soon Baik, In Young Yoon, Kun Hoon Kim, Jun Won Choi
Noisy Label Long Tailed Data Distributionally Robust Supervised Contrastive Loss Class Centroid Confidence Aware Contrastive

June 24, 2024

Model-Free Robust Reinforcement Learning with Sample Complexity Analysis
Yudan Wang, Shaofeng Zou, Yue Wang
Sample Complexity Model Free Monte Carlo Distributionally Robust

June 22, 2024

Distributionally Robust Constrained Reinforcement Learning under Strong Duality
Zhengfei Zhang, Kishan Panaganti, Laixi Shi, Yanan Sui, Adam Wierman, Yisong Yue
Convergence Guarantee Robust Reinforcement Learning Distributionally Robust Distributional Robustness Strong Duality Natural Distribution Shift

June 17, 2024

Statistical Learning of Distributionally Robust Stochastic Control in Continuous State Spaces
Shengbo Wang, Nian Si, Jose Blanchet, Zhengyuan Zhou
State Space Dynamic Programming Statistical Learning Distributionally Robust Robust Control Stochastic System

June 10, 2024

Robust Distribution Learning with Local and Global Adversarial Corruptions
Sloan Nietert, Ziv Goldfeld, Soroosh Shafiee
Robust Optimization Wasserstein Distance Sample Efficient Distributionally Robust DIstribution Estimation Adversarial Corruption Stochastic Robust Optimization

May 28, 2024

Sensor-Based Distributionally Robust Control for Safe Robot Navigation in Dynamic Environments
Kehan Long, Yinzhuang Yi, Zhirui Dai, Sylvia Herbert, Jorge Cortés, Nikolay Atanasov
Dynamic Environment Control Barrier Function Distributionally Robust Safe Robot Navigation Robust Navigation Uncertain Dynamic Environment

April 4, 2024

Distributionally Robust Reinforcement Learning with Interactive Data Collection: Fundamental Hardness and Near-Optimal Algorithm
Miao Lu, Han Zhong, Tong Zhang, Jose Blanchet
Robust Policy Robust Reinforcement Learning Hardness Result Distributionally Robust Near Optimal Algorithm

April 3, 2024

Distributionally Robust Policy and Lyapunov-Certificate Learning
Kehan Long, Jorge Cortes, Nikolay Atanasov
Lyapunov Function Neural Network Controller Distributionally Robust Chance Constraint Uncertain System Lyapunov Certificate

March 8, 2024

Integrating Predictive Motion Uncertainties with Distributionally Robust Risk-Aware Control for Safe Robot Navigation in Crowds
Kanghyun Ryu, Negar Mehr
Predictive Control Predictive Uncertainty Crowded Environment Distributionally Robust Human Trajectory Safe Robot Navigation Chance Constraint Chance Constrained

February 23, 2024

Distributionally Robust Off-Dynamics Reinforcement Learning: Provable Efficiency with Linear Function Approximation
Zhishuai Liu, Pan Xu
Linear Function Approximation Distributionally Robust Provable Benefit Online POMDP Solver Distributionally Robust Markov Decision Process

November 21, 2023

FedDRO: Federated Compositional Optimization for Distributionally Robust Learning
Prashant Khanduri, Chengyin Li, Rafi Ibn Sultan, Yao Qiang, Joerg Kliewer, Dongxiao Zhu
Federated Learning Distributionally Robust FedAvg Algorithm Compositional Optimization Compositional Operation

November 15, 2023

On the Foundation of Distributionally Robust Reinforcement Learning
Shengbo Wang, Nian Si, Jose Blanchet, Zhengyuan Zhou
Optimal Policy Flawed Foundation Robust Policy Robust Reinforcement Learning Distributionally Robust Hybrid Attack

October 27, 2023

Bridging Distributionally Robust Learning and Offline RL: An Approach to Mitigate Distribution Shift and Partial Data Coverage
Kishan Panaganti, Zaiyan Xu, Dileep Kalathil, Mohammad Ghavamzadeh
Offline Reinforcement Learning Constructive Approach Distribution Shift Limited Field Offline Reinforcement Learning Algorithm Distributionally Robust Distribution Shift Problem Single Policy Concentrability

October 21, 2023

Risk-Aware Wasserstein Distributionally Robust Control of Vessels in Natural Waterways
Juan Moreno Nadales, Astghik Hakobyan, David Muñoz de la Peña, Daniel Limon, Insoon Yang
Risk Aware Distributionally Robust Robust Controller Autonomous Surface Vessel Autonomous Vessel River Sandbank Maritime Industry

September 16, 2023

Distributionally Robust Post-hoc Classifiers under Prior Shifts
Jiaheng Wei, Harikrishna Narasimhan, Ehsan Amid, Wen-Sheng Chu, Yang Liu, Abhishek Kumar
Distribution Shift Test Time Distributionally Robust Two Shift Biased Training Robust Loss

Distributionally Robust

Papers

PSL: Rethinking and Improving Softmax Loss from Pairwise Perspective for Recommendation

Gradual Domain Adaptation via Manifold-Constrained Distributionally Robust Optimization

Upper and Lower Bounds for Distributionally Robust Off-Dynamics Reinforcement Learning

Distributionally Robust Inverse Reinforcement Learning for Identifying Multi-Agent Coordinated Sensing

Addressing Behavior Model Inaccuracies for Safe Motion Control in Uncertain Dynamic Environments

Distribution-Aware Robust Learning from Long-Tailed Data with Noisy Labels

Model-Free Robust Reinforcement Learning with Sample Complexity Analysis

Distributionally Robust Constrained Reinforcement Learning under Strong Duality

Statistical Learning of Distributionally Robust Stochastic Control in Continuous State Spaces

Robust Distribution Learning with Local and Global Adversarial Corruptions

Sensor-Based Distributionally Robust Control for Safe Robot Navigation in Dynamic Environments

Distributionally Robust Reinforcement Learning with Interactive Data Collection: Fundamental Hardness and Near-Optimal Algorithm

Distributionally Robust Policy and Lyapunov-Certificate Learning

Integrating Predictive Motion Uncertainties with Distributionally Robust Risk-Aware Control for Safe Robot Navigation in Crowds

Distributionally Robust Off-Dynamics Reinforcement Learning: Provable Efficiency with Linear Function Approximation

FedDRO: Federated Compositional Optimization for Distributionally Robust Learning

On the Foundation of Distributionally Robust Reinforcement Learning

Bridging Distributionally Robust Learning and Offline RL: An Approach to Mitigate Distribution Shift and Partial Data Coverage

Risk-Aware Wasserstein Distributionally Robust Control of Vessels in Natural Waterways

Distributionally Robust Post-hoc Classifiers under Prior Shifts