To circumvent this difficulty, we consider a new solution concept called robust agent policy, where agents aim to maximize the worst-case expected state value. We prove the existence of robust agent policy for finite state and finite action SAMGs.
Feb 9, 2024
Dec 6, 2022 · Abstract:Various methods for Multi-Agent Reinforcement Learning (MARL) have been developed with the assumption that agents' policies are ...
People also ask
How to implement multi-agent reinforcement learning?
Why do we need multi-agent reinforcement learning?
What is an example of state in reinforcement learning?
What is the difference between single agent and multi-agent reinforcement learning?
In this work, we propose a State-Adversarial Markov Game (SAMG) and make the first attempt to investigate different solution concepts of MARL under state ...
Our experiments show that adversarial state perturbations decrease agents' rewards for several baselines from the existing literature, while our algorithm ...
This work proposes a State-Adversarial Markov Game (SAMG) and makes the first attempt to investigate different solution concepts of MARL under state ...
In this work, we design a multi-agent reinforcement learning (MARL)-based framework for EAVs balancing in E-AMoD systems, with adversarial agents to model both ...
Apr 17, 2024 · Exciting news! Our paper titled "What is the Solution for State-Adversarial Multi-Agent Reinforcement Learning?" has been accepted for ...
This is the code for our paper "What is the Solution? A Robust Agent Policy for State Adversarial Multi-Agent Reinforcement Learning".
Abstract: Various methods for Multi-Agent Reinforcement Learning (MARL) have been developed with the assumption that agents' policies are based on accurate ...
What is the Solution for State-Adversarial Multi-Agent Reinforcement Learning? Songyang Han, Sanbao Su, Sihong He, and 4 more authors. Transactions on ...