HIER-~RCHICAL W~ULTI AGENT REINFORCEMENT LEARNING

منو

" HIER-~RCHICAL W~ULTI AGENT REINFORCEMENT LEARNING "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	150998
Doc. No	:	ET22790
Main Entry	:	Rajbala Makar
Title Proper	:	HIER-~RCHICAL W~ULTI AGENT REINFORCEMENT LEARNING
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	Reinforcement Learning has been extensively studied as a generalized approachfor sequential decision n~aking problems involving actions of a single agent. Hcnce.single agent learning and the challenges and issues involved are very well understood.Recent focus has been on co-operative multi agent learning as that opens up a wholenew dirncnsion to machine learning and the type of problems it can be used for. Thecomplesity introduced as a result of having a number of agents is a major deterrent forfornlulatiori of machine learning problems with multiple agents. Hierarchical learningmakes it easier to solve large problems b5- taking advantage of abstraction. This workfocuses on applying a hierarchical reinforcement learning approach to multi agent..,....-...,..tested for theQ1 PC1 bus cardBoth these projects mere sofixare des elopment efforts tonards contributing to dlfferentaspects of Roboucs and lZ1echatronics projects m the Controls and Roboucs Group..
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL46016.pdf
Title and statement of responsibility and	:	HIER-~RCHICAL W~ULTI AGENT REINFORCEMENT LEARNING [Thesis]