Home
Resource
Nonconvex representation learning from distributed datasets
PDF
PDF format is widely accepted and good for printing.
Plug-in required
PDF-1(1.85 MB)
Citation & Export
View Usage Statistics
Staff View

Citation & Export
Hide

Simple citation

Raja, Haroon. Nonconvex representation learning from distributed datasets. Retrieved from https://doi.org/doi:10.7282/t3-bm09-vj65

Export

Click here for information about Citation Management Tools at Rutgers.

Statistics
Hide

Description

TitleNonconvex representation learning from distributed datasets
NameRaja, Haroon (author); Bajwa, Waheed U. (chair); Gurbuzbalaban, Mert (internal member); Sarwate, Anand D. (internal member); Wakin, Michael B. (internal member); Yates, Roy D. (internal member); Rutgers University; School of Graduate Studies
Date Created2019
Other Date2019-10 (degree)
SubjectElectrical and Computer Engineering, Consensus averaging, Electronic data processing
Extent1 online resource (xi, 108 pages) : illustrations
DescriptionData representation is an important information processing task which finds use in diverse engineering applications like signal processing, machine learning, medical imaging, and geophysical data analysis, to name a few. Once a good representation model is known for a given application then the next question is learning that model under practical constraints imposed by the application. Two such constraints are i) data is available at geographically distributed sites, and ii) streaming data. In such scenarios distributed, decentralized, and online algorithms can be deployed for solving data representation problems, which are the focus of this dissertation. Specifically, this thesis focuses on solving following three problems: i) Solution for PCA for high-rate streaming data, ii) collaborative dictionary learning for big, distributed data, and iii) through the wall radar imaging (TWRI) in distributed settings. This thesis proposes new methods to tackle challenges arising due to distributed and steaming nature of data, providing a theoretical analysis of the proposed methodologies (except TWRI), and finally using simulations to demonstrate the efficacy of the proposed methods
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
Persistent URLhttps://doi.org/doi:10.7282/t3-bm09-vj65
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Version 8.5.5
Rutgers University Libraries - Copyright ©2024