Model Based Reinforcement Learning

Model-based reinforcement learning (MBRL) aims to improve the sample efficiency and robustness of reinforcement learning agents by learning a model of the environment's dynamics. Current research focuses on enhancing model accuracy and robustness through techniques like incorporating expert knowledge, using bisimulation metrics for state representation, and employing adversarial training to handle uncertainties. These advancements are leading to more efficient and reliable control policies in various applications, including robotics, autonomous driving, and even protein design, by reducing the need for extensive real-world interactions during training.

Papers

December 16, 2021

Sample-Efficient Reinforcement Learning via Conservative Model-Based Actor-Critic
Zhihai Wang, Jie Wang, Qi Zhou, Bin Li, Houqiang Li
Model Based Reinforcement Learning Model Based Sample Efficient Reinforcement Learning Free Counterpart Accurate Learning Conservative Objective Model

December 14, 2021

CEM-GD: Cross-Entropy Method with Gradient Descent Planner for Model-Based Reinforcement Learning
Kevin Huang, Sahin Lale, Ugo Rosolia, Yuanyuan Shi, Anima Anandkumar
Model Based Reinforcement Learning Cross Entropy Zeroth Order Optimal Action Stochastic Trajectory Trajectory Plan

December 9, 2021

An Experimental Design Perspective on Model-Based Reinforcement Learning
Viraj Mehta, Biswajit Paria, Jeff Schneider, Stefano Ermon, Willie Neiswanger
Reinforcement Learning Model Based Reinforcement Learning Experimental Design Data Efficient State Action Pair

December 8, 2021

Model-Value Inconsistency as a Signal for Epistemic Uncertainty
Angelos Filos, Eszter Vértes, Zita Marinho, Gregory Farquhar, Diana Borsa, Abram Friesen, Feryal Behbahani, Tom Schaul, André Barreto, Simon Osindero
Model Based Reinforcement Learning Epistemic Uncertainty Value Function Abnormal Signal Value Based Consistent Model

December 6, 2021

ED2: Environment Dynamics Decomposition World Models for Continuous Control
Jianye Hao, Yifu Yuan, Cong Wang, Zhen Wang
Model Based Reinforcement Learning World Model Model Free Continuous Control Dynamic Model Environment Dynamic

December 2, 2021

Maximum Entropy Model-based Reinforcement Learning
Oleg Svidchenko, Aleksei Shpilman
Reinforcement Learning Model Based Reinforcement Learning Maximum Entropy Reinforcement Learning Agent Environment Interaction

November 16, 2021

On Effective Scheduling of Model-based Reinforcement Learning
Hang Lai, Jian Shen, Weinan Zhang, Yimin Huang, Xing Zhang, Ruiming Tang, Yong Yu, Zhenguo Li
Real Time Model Based Reinforcement Learning Policy OpTimization Efficient Scheduling Model Based Policy

November 14, 2021

Free Will Belief as a consequence of Model-based Reinforcement Learning
Erik M. Rehn
Reinforcement Learning LeArning Abstract Model Based Reinforcement Learning Human Decision Artificial Intelligence Agent Human Learning Experiential Consequence

November 11, 2021

Model-Based Reinforcement Learning via Stochastic Hybrid Models
Hany Abdulsamad, Jan Peters
Model Based Reinforcement Learning Nonlinear Dynamic Nonlinear Model Hybrid Modeling Piecewise Polynomial

November 9, 2021

Risk Sensitive Model-Based Reinforcement Learning using Uncertainty Guided Planning
Stefan Radic Webster, Peter Flach
Model Based Reinforcement Learning Epistemic Uncertainty Risk Sensitive Uncertainty Aware Uncertain Region