Exogenous Global Markov Process

Exogenous global Markov processes model scenarios where an unobservable, globally influencing process affects multiple independent systems or agents. Current research focuses on developing algorithms, such as LEMP, to learn optimal strategies in these settings, particularly within the framework of reinforcement learning and restless multi-armed bandits, aiming to mitigate the impact of this exogenous influence on decision-making. This research is significant because it addresses the challenge of learning and decision-making under uncertainty and uncontrolled external factors, with applications in areas like resource allocation and adaptive control systems. Improved algorithms offer the potential for more efficient and robust solutions in complex, real-world problems.

Papers

May 29, 2024

Convergence Bounds for Sequential Monte Carlo on Multimodal Distributions using Soft Decomposition
Holden Lee, Matheau Santana-Gijzen
Markov Chain Monte Carlo Microbial Decomposition Sequential Monte Carlo Multimodal Distribution Mixing Time Convergence Bound Exogenous Global Markov Process

July 1, 2023

Causal Structure Learning by Using Intersection of Markov Blankets
Yiran Dong, Chuanhou Gao
Structural Causal Model Bayesian Network Intersection Scenario Causal Structure Learning Markov Blanket Exogenous Global Markov Process

March 22, 2023

Reinforcement Learning with Exogenous States and Rewards
George Trimponias, Thomas G. Dietterich
Reinforcement Learning Markov Decision Process State Space Reward Report MDP Model Exogenous Global Markov Process

June 9, 2022

Sample-Efficient Reinforcement Learning in the Presence of Exogenous Information
Yonathan Efroni, Dylan J. Foster, Dipendra Misra, Akshay Krishnamurthy, John Langford
Reinforcement Learning Speech Presence Sample Efficient Reinforcement Learning Exogenous Information Exogenous Global Markov Process

February 28, 2022

Restless Multi-Armed Bandits under Exogenous Global Markov Process
Tomer Gafni, Michal Yemini, Kobi Cohen
Markov Decision Process Multi Armed Bandit Restless Multi Armed Bandit Logarithmic Regret Dynamic Restless Exogenous Global Markov Process

December 17, 2021

Learning in Restless Bandits under Exogenous Global Markov Process
Tomer Gafni, Michal Yemini, Kobi Cohen
LeArning Abstract Markov Decision Process Restless Multi Armed Bandit Restless Bandit Logarithmic Regret Exogenous Global Markov Process