Robust Deep Reinforcement Learning

Robust deep reinforcement learning (DRL) aims to create AI agents capable of performing reliably in unpredictable environments, addressing the fragility of standard DRL methods to variations in conditions or adversarial attacks. Current research focuses on enhancing robustness through techniques like adversarial training, risk-sensitive algorithms (e.g., employing exponential criteria or quantile regression), and adaptive perturbation methods that dynamically adjust training difficulty. These advancements are crucial for deploying DRL in real-world applications, such as autonomous driving and robotics, where safety and reliability are paramount, and for improving the generalizability and trustworthiness of DRL models.

Papers

October 15, 2024

Learning Agents With Prioritization and Parameter Noise in Continuous State and Action Space
Rajesh Mangannavar, Gopalakrishnan Srinivasaraghavan
Action Space Deep Q Learning Deep Deterministic Policy Gradient Continuous Speech Learning Agent Priority Based Noise Modeling Robust Deep Reinforcement Learning

August 13, 2024

Robust Deep Reinforcement Learning for Inverter-based Volt-Var Control in Partially Observable Distribution Networks
Qiong Liu, Ye Guo, Tong Xu
Distribution System DRL Algorithm Robust Deep Reinforcement Learning Active Distribution Based Volt Var Control

June 23, 2024

Understanding and Diagnosing Deep Reinforcement Learning
Ezgi Korkmaz
Deep Learning Deep Reinforcement Learning Human Understanding Decision Boundary Robust Deep Reinforcement Learning

June 12, 2024

Time-Constrained Robust MDPs
Adil Zouitine, David Bertoin, Pierre Clavier, Matthieu Geist, Emmanuel Rachelson
Robust Reinforcement Learning Robust Deep Reinforcement Learning Time Constraint

May 20, 2024

Robust Deep Reinforcement Learning with Adaptive Adversarial Perturbations in Action Space
Qianmei Liu, Yufei Kuang, Jie Wang
Deep Reinforcement Learning Adversarial Perturbation Adversarial Learning Action Space Robust Deep Reinforcement Learning

March 1, 2024

Robust Deep Reinforcement Learning Through Adversarial Attacks and Training : A Survey
Lucas Schott, Josephine Delas, Hatem Hajri, Elies Gherbi, Reda Yaich, Nora Boulahia-Cuppens, Frederic Cuppens, Sylvain Lamprier
Timely Survey Adversarial Attack Deep Reinforcement Learning Training Data Adversarial Training Robust Deep Reinforcement Learning

February 15, 2024

Risk-Sensitive Soft Actor-Critic for Robust Deep Reinforcement Learning under Distribution Shifts
Tobias Enders, James Harrison, Maximilian Schiffer
Distribution Shift Risk Sensitive Robust Policy Soft Actor Critic Risk Sensitive Reinforcement Learning Robust Deep Reinforcement Learning Risk Sensitive Algorithm

January 4, 2024

A Survey Analyzing Generalization in Deep Reinforcement Learning
Ezgi Korkmaz
Timely Survey Deep Reinforcement Learning Strong Generalization Reinforcement Learning Research Robust Deep Reinforcement Learning

January 1, 2024

Adversarially Trained Weighted Actor-Critic for Safe Offline Reinforcement Learning
Honghao Wei, Xiyue Peng, Arnob Ghosh, Xin Liu
Offline Reinforcement Learning Robust Deep Reinforcement Learning Offline Safe Reinforcement Learning

April 25, 2023

A Multi-Task Approach to Robust Deep Reinforcement Learning for Resource Allocation
Steffen Gracla, Carsten Bockelmann, Armin Dekorsy
Multi Task Learning Multi Task Resource Allocation Actor Critic Algorithm Robust Deep Reinforcement Learning Gradient Episodic Memory

April 14, 2023

A Platform-Agnostic Deep Reinforcement Learning Framework for Effective Sim2Real Transfer towards Autonomous Driving
Dianzhao Li, Ostap Okhrin
Deep Reinforcement Learning Autonomous Driving Perception Module Robust Deep Reinforcement Learning Sim2Real Transfer

February 14, 2023

Regret-Based Defense in Adversarial Reinforcement Learning
Roman Belaire, Pradeep Varakantham, Thanh Nguyen, David Lo
Reinforcement Learning Deep Reinforcement Learning Adversarial Example Adversarial Reinforcement Learning Robust Deep Reinforcement Learning

December 18, 2022

Risk-Sensitive Reinforcement Learning with Exponential Criteria
Erfaun Noorani, Christos Mavridis, John Baras
Reinforcement Learning Actor Critic Algorithm Risk Sensitive Reinforcement Learning Robust Deep Reinforcement Learning

June 21, 2022

Robust Deep Reinforcement Learning through Bootstrapped Opportunistic Curriculum
Junlin Wu, Yevgeniy Vorobeychik
Adversarial Robustness Adversarial Perturbation Robust Reinforcement Learning Robust Deep Reinforcement Learning Adversarial Curriculum

November 13, 2021

Robust Deep Reinforcement Learning for Extractive Legal Summarization
Duy-Hung Nguyen, Bao-Sinh Nguyen, Nguyen Viet Dung Nghiem, Dung Tien Le, Mim Amina Khatun, Minh-Tien Nguyen, Hung Le
Reinforcement Learning Automatic Summarization Legal Datasets Robust Deep Reinforcement Learning Legal Summarization

November 6, 2021

Robust Deep Reinforcement Learning for Quadcopter Control
Aditya M. Deshpande, Ali A. Minai, Manish Kumar
Deep Reinforcement Learning Robust Policy Quadrotor Control Drone Control Robust Deep Reinforcement Learning

Robust Deep Reinforcement Learning

Papers

Learning Agents With Prioritization and Parameter Noise in Continuous State and Action Space

Robust Deep Reinforcement Learning for Inverter-based Volt-Var Control in Partially Observable Distribution Networks

Understanding and Diagnosing Deep Reinforcement Learning

Time-Constrained Robust MDPs

Robust Deep Reinforcement Learning with Adaptive Adversarial Perturbations in Action Space

Robust Deep Reinforcement Learning Through Adversarial Attacks and Training : A Survey

Risk-Sensitive Soft Actor-Critic for Robust Deep Reinforcement Learning under Distribution Shifts

A Survey Analyzing Generalization in Deep Reinforcement Learning

Adversarially Trained Weighted Actor-Critic for Safe Offline Reinforcement Learning

A Multi-Task Approach to Robust Deep Reinforcement Learning for Resource Allocation

A Platform-Agnostic Deep Reinforcement Learning Framework for Effective Sim2Real Transfer towards Autonomous Driving

Regret-Based Defense in Adversarial Reinforcement Learning

Risk-Sensitive Reinforcement Learning with Exponential Criteria

Robust Deep Reinforcement Learning through Bootstrapped Opportunistic Curriculum

Robust Deep Reinforcement Learning for Extractive Legal Summarization

Robust Deep Reinforcement Learning for Quadcopter Control