A General Approach to Signal Decomposition Using Reinforcement Learning with Custom Signal Dictionaries

دانشگاه آزاد اسلامی اصفهان(خوراسگان)

منو

رکورد قبلی

رکورد بعدی

" A General Approach to Signal Decomposition Using Reinforcement Learning with Custom Signal Dictionaries "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	151557
Doc. No	:	ET23349
Main Entry	:	Nigel A. Ziyad
Title Proper	:	A General Approach to Signal Decomposition Using Reinforcement Learning with Custom Signal Dictionaries
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	Herein describes a proposed method to estimate signaIs by an adaptivedecomposition. These signals rtre likely to be nonstationary signds, that is, varying infrequency over time, so it is proposed that a scheme using a set of known functions, referredto as "the dictionary," can be used and adaptively improved by a learning method. Thisscheme allows for greater flexibility in applications that limit the ability for an orthogonalsolution or that require a non-orthogonal description of the signal under analysis becausethe a priori information is incomplete. Other methods that use dictionaries are generallylimited to the type of dictionaries that are valid for the method. A general and adaptivemethod would allow ao flexibility of dictionary selection, and bt the possibility ofclassification and/or compression. Thus, this work makes three main contributions:1e It describes a unique philosophy and proof-ofconcept for utilizing customizabledictionaries: a mechanism to allow user heuristics.2p It demonstrates the possibility of such a mechanism, using a simple learning algorithmand criteria.3a It stimulates further interest in this area of work through.....-....,....-...,..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	:	TL46588.pdf
Title and statement of responsibility and	:	A General Approach to Signal Decomposition Using Reinforcement Learning with Custom Signal Dictionaries [Thesis]