Partially Observed Markov Decision Process

Partially Observed Markov Decision Processes (POMDPs) model decision-making under uncertainty, where the complete state of the environment is unknown. Current research focuses on developing efficient algorithms, such as those incorporating options, to learn optimal policies in these complex scenarios, often leveraging techniques like point-based methods or function approximation within model architectures like actor-critic networks. This work is significant because it addresses the challenges of learning in situations with incomplete information, with applications ranging from robotics and dialogue systems to medical imaging and financial trading. Improved POMDP solutions promise more robust and adaptable AI systems across diverse fields.

Papers

July 26, 2024

SOAP-RL: Sequential Option Advantage Propagation for Reinforcement Learning in POMDP Environments
Shu Ishida, João F. Henriques
Reinforcement Learning Markov Decision Process POMDP Policy Partially Observed Markov Decision Process Option Learning

October 11, 2023

Prospective Side Information for Latent MDPs
Jeongyeol Kwon, Yonathan Efroni, Shie Mannor, Constantine Caramanis
Markov Decision Process O$ Regret Near Optimal Policy Side Information Partially Observed Markov Decision Process Latent MDPs Latent Markov Decision Process

June 27, 2023

An analytical model of active inference in the Iterated Prisoner's Dilemma
Daphne Demekas, Conor Heins, Brennan Klein
Game Theoretic Active Inference Mathematical Model Emotional Dilemma Partially Observed Markov Decision Process Bayesian Agent

April 10, 2023

A Novel Point-based Algorithm for Multi-agent Control Using the Common Information Approach
Dengwang Tang, Ashutosh Nayyar, Rahul Jain
Practical Algorithm Markov Decision Process Multi Agent Long Form Novel Information Approach Partially Observed Markov Decision Process

February 19, 2023

Generalization in Visual Reinforcement Learning with the Reward Sequence Distribution
Jie Wang, Rui Yang, Zijie Geng, Zhihao Shi, Mingxuan Ye, Qi Zhou, Shuiwang Ji, Bin Li, Yongdong Zhang, Feng Wu
Strong Generalization Markov Decision Process State Space Visual Reinforcement Learning Partially Observed Markov Decision Process Reward Distribution

December 5, 2022

L2SR: Learning to Sample and Reconstruct for Accelerated MRI via Reinforcement Learning
Pu Yang, Bin Dong
Reinforcement Learning Joint Learning External Sample Accelerated Magnetic Resonance Imaging Partially Observed Markov Decision Process Reconstruction Learning

July 5, 2022

Deep Reinforcement Learning Approach for Trading Automation in The Stock Market
Taylan Kabbani, Ekrem Duman
Deep Reinforcement Learning Markov Decision Process Human Decision Deep Reinforcement Learning Approach Stock Market Delayed Deep Deterministic Policy Gradient Partially Observed Markov Decision Process Automated Cryptocurrency Trading

May 26, 2022

Embed to Control Partially Observed Systems: Representation Learning with Provable Sample Efficiency
Lingxiao Wang, Qi Cai, Zhuoran Yang, Zhaoran Wang
Reinforcement Learning Representation Learning Markov Decision Process Reinforcement Learning Algorithm Policy OpTimization Sample Efficiency Partially Observed Partially Observed Markov Decision Process

April 20, 2022

Reinforcement Learning from Partial Observation: Linear Function Approximation with Provable Sample Efficiency
Qi Cai, Zhuoran Yang, Zhaoran Wang
Reinforcement Learning Markov Decision Process Optimal Policy Linear Function Approximation Sample Efficiency Partial Observability Partial Observation Partially Observed Markov Decision Process

February 20, 2022

Finite-Time Analysis of Natural Actor-Critic for POMDPs
Semih Cayci, Niao He, R. Srikant
Markov Decision Process Finite Time Reinforcement Learning Problem Partially Observed Markov Decision Process Finite State Controller

December 22, 2021

Entropy-Regularized Partially Observed Markov Decision Processes
Timothy L. Molloy, Girish N. Nair
Markov Decision Process State Estimation POMDP Solver Partially Observed Markov Decision Process

Partially Observed Markov Decision Process

Papers

SOAP-RL: Sequential Option Advantage Propagation for Reinforcement Learning in POMDP Environments

Prospective Side Information for Latent MDPs

An analytical model of active inference in the Iterated Prisoner's Dilemma

A Novel Point-based Algorithm for Multi-agent Control Using the Common Information Approach

Generalization in Visual Reinforcement Learning with the Reward Sequence Distribution

L2SR: Learning to Sample and Reconstruct for Accelerated MRI via Reinforcement Learning

Deep Reinforcement Learning Approach for Trading Automation in The Stock Market

Embed to Control Partially Observed Systems: Representation Learning with Provable Sample Efficiency

Reinforcement Learning from Partial Observation: Linear Function Approximation with Provable Sample Efficiency

Finite-Time Analysis of Natural Actor-Critic for POMDPs

Entropy-Regularized Partially Observed Markov Decision Processes