Modeling Others using Oneself in Multi-Agent Reinforcement Learning

02/26/2018 ∙ by Roberta Raileanu, et al. ∙ 0 ∙ share

We consider the multi-agent reinforcement learning setting with imperfect information in which each agent is trying to maximize its own utility. The reward function depends on the hidden state (or goal) of both agents, so the agents must infer the other players' hidden goals from their observed behavior in order to solve the tasks. We propose a new approach for learning in these domains: Self Other-Modeling (SOM), in which an agent uses its own policy to predict the other agent's actions and update its belief of their hidden state in an online manner. We evaluate this approach on three different tasks and show that the agents are able to learn better policies using their estimate of the other players' hidden states, in both cooperative and adversarial settings.

READ FULL TEXT

Universal Policies to Learn Them All

We explore a collaborative and cooperative multi-agent reinforcement lea...
08/24/2019 ∙ by Hassam Ullah Sheikh, et al. ∙ 0 ∙ share

read it
Feudal Multi-Agent Hierarchies for Cooperative Reinforcement Learning

We investigate how reinforcement learning agents can learn to cooperate....
01/24/2019 ∙ by Sanjeevan Ahilan, et al. ∙ 0 ∙ share

read it
Prosocial or Selfish? Agents with different behaviors for Contract Negotiation using Reinforcement Learning

We present an effective technique for training deep learning agents capa...
09/19/2018 ∙ by Vishal Sunder, et al. ∙ 0 ∙ share

read it
Robust Opponent Modeling via Adversarial Ensemble Reinforcement Learning in Asymmetric Imperfect-Information Games

This paper presents an algorithmic framework for learning robust policie...
09/18/2019 ∙ by Macheng Shen, et al. ∙ 0 ∙ share

read it
Computational Rationalization: The Inverse Equilibrium Problem

Modeling the purposeful behavior of imperfect agents from a small number...
08/15/2013 ∙ by Kevin Waugh, et al. ∙ 0 ∙ share

read it
Do deep reinforcement learning agents model intentions?

Inferring other agents' mental states such as their knowledge, beliefs a...
05/15/2018 ∙ by Tambet Matiisen, et al. ∙ 0 ∙ share

read it
Learning Multi-agent Implicit Communication Through Actions: A Case Study in Contract Bridge, a Collaborative Imperfect-Information Game

In situations where explicit communication is limited, a human collabora...
10/10/2018 ∙ by Zheng Tian, et al. ∙ 10 ∙ share

read it

Modeling Others using Oneself in Multi-Agent Reinforcement Learning

Authors

Universal Policies to Learn Them All

Feudal Multi-Agent Hierarchies for Cooperative Reinforcement Learning

Prosocial or Selfish? Agents with different behaviors for Contract Negotiation using Reinforcement Learning

Robust Opponent Modeling via Adversarial Ensemble Reinforcement Learning in Asymmetric Imperfect-Information Games

Computational Rationalization: The Inverse Equilibrium Problem

Do deep reinforcement learning agents model intentions?

Learning Multi-agent Implicit Communication Through Actions: A Case Study in Contract Bridge, a Collaborative Imperfect-Information Game

This week in AI