Markov Reward Process

Markov Reward Processes (MRPs) model sequential decision-making problems where an agent receives rewards based on its actions and the resulting state transitions. Current research focuses on improving the efficiency and robustness of algorithms for solving MRP-based problems, particularly in reinforcement learning contexts, including developing novel exploration strategies, addressing limitations in offline imitation learning, and enhancing the convergence and sample efficiency of algorithms like Temporal Difference learning and Natural Policy Gradient methods. These advancements are crucial for tackling complex real-world applications, such as resource allocation, robotics control, and even supervised learning tasks where data dependencies are better modeled through MRP frameworks.

Papers

August 18, 2024

Directed Exploration in Reinforcement Learning from Linear Temporal Logic
Marco Bagatella, Andreas Krause, Georg Martius
Reinforcement Learning Temporal Logic Directed Exploration Temporal Logic Formula Markov Reward Process

August 17, 2024

Markov Balance Satisfaction Improves Performance in Strictly Batch Offline Imitation Learning
Rishabh Agrawal, Nathan Dahlin, Rahul Jain, Ashutosh Nayyar
Imitation Learning System Performance Based Imitation Offline Imitation Learning Transition Dynamic Markov Reward Process

June 12, 2024

April 23, 2024

An MRP Formulation for Supervised Learning: Generalized Temporal Difference Learning Models
Yangchen Pan, Junfeng Wen, Chenjun Xiao, Philip Torr
Reinforcement Learning Supervised Learning Temporal Difference Learning Temporal Difference Markov Reward Process

October 18, 2023

Improved Sample Complexity Analysis of Natural Policy Gradient Algorithm with General Parameterization for Infinite Horizon Discounted Reward Markov Decision Processes
Washim Uddin Mondal, Vaneet Aggarwal
Markov Decision Process Natural Policy Gradient Infinite Horizon Accelerated Gradient Descent Markov Reward Process

April 11, 2023

A Tale of Sampling and Estimation in Discounted Reinforcement Learning
Alberto Maria Metelli, Mirco Mutti, Marcello Restelli
Reinforcement Learning Estimation Task Policy OpTimization Open Sampling Episodic Reinforcement Learning Cautionary TALE Markov Reward Process Moving Horizon

April 7, 2023

Full Gradient Deep Reinforcement Learning for Average-Reward Criterion
Tejas Pagare, Vivek Borkar, Konstantin Avrachenkov
DQN Agent Restless Multi Armed Bandit Average Reward Markov Reward Process

April 3, 2023

A Tutorial Introduction to Reinforcement Learning
Mathukumalli Vidyasagar
Reinforcement Learning Markov Decision Process Tutorial Review Stochastic Approximation Q$ Learning Markov Reward Process

November 29, 2022

Performance Evaluation, Optimization and Dynamic Decision in Blockchain Systems: A Recent Overview
Quan-Lin Li, Yan-Xia Chang, Qing Wang
Optimization Purpose Markov Decision Process Performance Evaluation Blockchain Technology Recent Development Blockchain System Dynamic Decision Markov Reward Process

November 7, 2022

Policy evaluation from a single path: Multi-step methods, mixing and mis-specification
Yaqi Duan, Martin J. Wainwright
Policy Evaluation Value Function Multi Step Non Asymptotic Design Specification Mixing Process Markov Reward Process

June 27, 2022

Utility Theory for Sequential Decision Making
Mehran Shakerinava, Siamak Ravanbakhsh
Reinforcement Learning Markov Decision Process Sequential Decision Sequential Decision Making Utility Based Markov Reward Process

Markov Reward Process

Papers

Directed Exploration in Reinforcement Learning from Linear Temporal Logic

Markov Balance Satisfaction Improves Performance in Strictly Batch Offline Imitation Learning

A Federated Online Restless Bandit Framework for Cooperative Resource Allocation

Finite Time Analysis of Temporal Difference Learning for Mean-Variance in a Discounted MDP

An MRP Formulation for Supervised Learning: Generalized Temporal Difference Learning Models

Improved Sample Complexity Analysis of Natural Policy Gradient Algorithm with General Parameterization for Infinite Horizon Discounted Reward Markov Decision Processes

A Tale of Sampling and Estimation in Discounted Reinforcement Learning

Full Gradient Deep Reinforcement Learning for Average-Reward Criterion

A Tutorial Introduction to Reinforcement Learning

Performance Evaluation, Optimization and Dynamic Decision in Blockchain Systems: A Recent Overview

Policy evaluation from a single path: Multi-step methods, mixing and mis-specification

Utility Theory for Sequential Decision Making