Reinforcement Learning Performance

Reinforcement learning (RL) research aims to improve the performance and robustness of algorithms that learn optimal behaviors through trial and error. Current efforts focus on addressing challenges like distribution shifts during testing, understanding why deep RL methods succeed despite using random exploration, and developing more efficient algorithms like those inspired by value iteration. These advancements are crucial for building reliable RL agents applicable to diverse real-world problems, from manufacturing optimization to personalized user experiences, and for establishing more rigorous evaluation methodologies.

Papers

February 5, 2024

Assessing the Impact of Distribution Shift on Reinforcement Learning Performance
Ted Fujimoto, Joshua Suetterlein, Samrat Chatterjee, Auroop Ganguly
Reinforcement Learning Global Impact Distribution Shift Reinforcement Learning Algorithm Reinforcement Learning Performance

December 13, 2023

The Effective Horizon Explains Deep RL Performance in Stochastic Environments
Cassidy Laidlaw, Banghua Zhu, Stuart Russell, Anca Dragan
Reinforcement Learning Deep Reinforcement Learning DQN Agent Stochastic Environment Exploration Algorithm Reinforcement Learning Performance

August 9, 2023

Variations on the Reinforcement Learning performance of Blackjack
Avish Buramdoyal, Tim Gebbie
Category Wise Variation Convergence Behavior Poker Hand Reinforcement Learning Performance

April 13, 2023

Deep reinforcement learning applied to an assembly sequence planning problem with user preferences
Miguel Neves, Pedro Neto
Deep Reinforcement Learning Q Learning Deep Q Learning User Preference Assembly Planning Tabular Q Learning Reinforcement Learning Performance

September 29, 2022

Does Zero-Shot Reinforcement Learning Exist?
Ahmed Touati, Jérémy Rapin, Yann Ollivier
Zero Shot Reinforcement Learning Performance

August 4, 2022

Human Decision Makings on Curriculum Reinforcement Learning with Difficulty Adjustment
Yilei Zeng, Jiali Duan, Yang Li, Emilio Ferrara, Lerrel Pinto, C. -C. Jay Kuo, Stefanos Nikolaidis
Reinforcement Learning Human Decision Human Centered Curriculum Reinforcement Learning Large Scale Reinforcement Learning Reinforcement Learning Performance Difficulty Adjustment