Learning without Gradients: Multi-Agent Reinforcement Learning approach to optimization

Amir Morcos, Hong Man, Aaron West, Brian Maguire

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The field of Reinforcement Learning continues to show promise in solving old problems in new innovative ways. Thanks to the algorithms' ability to learn without an explicit set of labeled training data, the action, environment, reward approach has lured many researches into framing old problems in this manner. Recent publications have demonstrated how utilizing a multi-agent reinforcement learning approach can lead to a superior policy for optimization algorithm over the current standards. The challenge with the aforementioned approaches is the inclusion of the gradient in the state-space. This forces a costly calculation that is often the bottle neck in most machine learning problems, often limiting or preventing training at the edge or on the front lines. While previous works dating back decades have demonstrated the ability to train simple machine learning models without the use of gradients, none have done so using a policy which leverages previous experiences to solve the problem more quickly. This work will show how a Multi-Agent Reinforcement Learning approach can be used to optimize models in training without the need for the gradient of the loss function, effectively eliminating the need for backpropagation and significantly reducing the computational power required to train a model. Furthermore, the work will examine conditions under which the agents failed to find an optimal solution. As well as how this approach can be beneficial in complex defense applications.

Original languageEnglish
Title of host publicationArtificial Intelligence and Machine Learning in Defense Applications IV
EditorsJudith Dijk
ISBN (Electronic)9781510655553
DOIs
StatePublished - 2022
EventArtificial Intelligence and Machine Learning in Defense Applications IV 2022 - Berlin, Germany
Duration: 6 Sep 20227 Sep 2022

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume12276
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceArtificial Intelligence and Machine Learning in Defense Applications IV 2022
Country/TerritoryGermany
CityBerlin
Period6/09/227/09/22

Fingerprint

Dive into the research topics of 'Learning without Gradients: Multi-Agent Reinforcement Learning approach to optimization'. Together they form a unique fingerprint.

Cite this