Modern Reinforcement Learning

Modern reinforcement learning (RL) aims to develop algorithms enabling agents to learn optimal behaviors through trial-and-error interaction with an environment, focusing on efficient exploration and robust policy selection. Current research emphasizes improving exploration strategies, including Bayesian methods and causal inference techniques, as well as addressing limitations of the reward hypothesis and developing more efficient algorithms like accelerated value iteration and novel loss functions based on the distribution of Bellman errors. These advancements are driving progress in diverse applications, such as robotics, autonomous systems, and policy analysis, by enabling more efficient learning and better generalization to real-world scenarios.

Papers

October 17, 2024

Deep Reinforcement Learning for Online Optimal Execution Strategies
Alessandro Micheli, Mélodie Monod
Deep Reinforcement Learning Deep Deterministic Policy Gradient Non Markovian Modern Reinforcement Learning Optimal Execution

August 19, 2024

Efficient Exploration in Deep Reinforcement Learning: A Novel Bayesian Actor-Critic Algorithm
Nikolai Rozanov
Reinforcement Learning Deep Reinforcement Learning Efficient Exploration Modern Reinforcement Learning

July 17, 2024

Variable-Agnostic Causal Exploration for Reinforcement Learning
Minh Hoang Nguyen, Hung Le, Svetha Venkatesh
Reinforcement Learning Causal Discovery Modern Reinforcement Learning Agent Exploration Causal Exploration

July 15, 2024

Three Dogmas of Reinforcement Learning
David Abel, Mark K. Ho, Anna Harutyunyan
Reinforcement Learning Intelligent Agent Reward Signal Modern Reinforcement Learning

July 11, 2024

A Deep Reinforcement Learning Framework and Methodology for Reducing the Sim-to-Real Gap in ASV Navigation
Luis F W Batista, Junghwan Ro, Antoine Richard, Pete Schroepfer, Seth Hutchinson, Cedric Pradalier
Deep Reinforcement Learning NCD Method Autonomous Surface Vehicle Real World Experiment Modern Reinforcement Learning

January 21, 2024

Emergent Dominance Hierarchies in Reinforcement Learning Agents
Ram Rachum, Yonatan Nakar, Bill Tomlinson, Nitay Alon, Reuth Mirsky
Multi Agent Reinforcement Learning Reinforcement Learning Agent Intrinsic Reward Dominance Hierarchy Modern Reinforcement Learning

December 11, 2023

Can Reinforcement Learning support policy makers? A preliminary study with Integrated Assessment Models
Theodore Wolf, Nantas Nardelli, John Shawe-Taylor, Maria Perez-Ortiz
Preliminary Study Based Policy Modern Reinforcement Learning Public Policy Policy Making

November 14, 2023

Deep Reinforcement Learning for 2D Physics-Based Object Manipulation in Clutter
Luca Renna
Deep Reinforcement Learning Markov Decision Process Policy Gradient 2 Dimensional Reinforcement Learning Algorithm Dense Clutter Modern Reinforcement Learning

July 5, 2023

LLQL: Logistic Likelihood Q-Learning for Reinforcement Learning
Outongyi Lv, Bingxin Zhou
Reinforcement Learning Bellman Equation Modern Reinforcement Learning Bellman Error Logistic Model

June 12, 2023

Composing Efficient, Robust Tests for Policy Selection
Dustin Morrill, Thomas J. Walsh, Daniel Hernandez, Peter R. Wurman, Peter Stone
Test Case Modern Reinforcement Learning Poker Hand Text Editing High Performing Policy Robust Test Policy Selection

May 26, 2023

Accelerating Value Iteration with Anchoring
Jongmin Lee, Ernest K. Ryu
Value Iteration Modern Reinforcement Learning Bellman Error Optimality Criterion Independent Rate

February 22, 2023

Provably Efficient Reinforcement Learning via Surprise Bound
Hanlin Zhu, Ruosong Wang, Jason D. Lee
Value Function Efficient Reinforcement Learning Modern Reinforcement Learning Surprise Bound

February 20, 2023

Reinforcement Learning with Function Approximation: From Linear to Nonlinear
Jihao Long, Jiequn Han
Reinforcement Learning Nonlinear Model Function Approximation Linear Ordered Data Modern Reinforcement Learning Nonlinear Function Approximation

February 15, 2023

TiZero: Mastering Multi-Agent Football with Curriculum Learning and Self-Play
Fanqi Lin, Shiyu Huang, Tim Pearce, Wenze Chen, Wei-Wei Tu
Curriculum Learning Self Play Multi Agent Coordination Multi Agent Particle Modern Reinforcement Learning

February 1, 2023

Sample Complexity of Kernel-Based Q-Learning
Sing-Yuan Yeh, Fu-Chieh Chang, Chang-Wei Yueh, Pei-Yuan Wu, Alberto Bernacchia, Sattar Vakili
Sample Complexity Q Learning Optimal Policy Modern Reinforcement Learning Kernel Based Reinforcement Learning

September 23, 2022

Unified Algorithms for RL with Decision-Estimation Coefficients: PAC, Reward-Free, Preference-Based Learning, and Beyond
Fan Chen, Song Mei, Yu Bai
Sample Complexity Bound Reward Free Modern Reinforcement Learning Unified Algorithm

September 5, 2022

Natural Policy Gradients In Reinforcement Learning Explained
W. J. A. van Heeswijk
Reinforcement Learning Policy Gradient Proximal Policy Optimization Natural Policy Gradient Trust Region Policy Optimization Modern Reinforcement Learning

June 25, 2022

Guided Exploration in Reinforcement Learning via Monte Carlo Critic Optimization
Igor Kuznetsov
Reinforcement Learning Exploration Method Guided Exploration Modern Reinforcement Learning Policy Deep Reinforcement Learning Action Correction Unsupervised Calibration

February 23, 2022

Globally Convergent Policy Search over Dynamic Filters for Output Estimation
Jack Umenberger, Max Simchowitz, Juan C. Perdomo, Kaiqing Zhang, Russ Tedrake
Lyapunov Function Partial Observability Modern Reinforcement Learning Direct Policy Dynamic Filter

December 1, 2021

Wish you were here: Hindsight Goal Selection for long-horizon dexterous manipulation
Todor Davchev, Oleg Sushkov, Jean-Baptiste Regli, Stefan Schaal, Yusuf Aytar, Markus Wulfmeier, Jon Scholz
Sparse Reward Long Horizon Task Long Horizon Manipulation Task Modern Reinforcement Learning Agent Exploration