Reinforcement Learning in Decentralized Stochastic Control Systems with Partial History Sharing

Published on 2015-07-282166 Views

Jalal Arabneydi

In this paper, we are interested in systems with multiple agents that wish to cooperate in order to accomplish a common task while a) agents have different information (decentralized information) and

RLDM 2015 - Edmonton

Related categories

Presentation

Reinforcement Learning in Multi-Agent Systems with Partial History Sharing00:00

Motivation00:25

Challenges01:12

Problem Formulation - 101:44

Problem Formulation - 202:19

Partial History Sharing03:11

Methodology04:38

Salient Feature of the Approach05:11

Step 1) Common Information Approach - 106:05

Step 1) Common Information Approach - 206:57

Step 2: An Approximate POMDP RL Algorithm - 107:31

Step 2: An Approximate POMDP RL Algorithm- 208:54

Multi-Agent RL Algorithm10:22

Example: Multi-Access Broadcasting Channel (MABC) - 112:16

Example: Multi-Access Broadcasting Channel (MABC) - 213:29

Example: Multi-Access Broadcasting Channel (MABC) - 314:05

Summary14:57

Thank you15:39