INDIGO Home University of Illinois at Urbana-Champaign logo uic building uic pavilion uic student center

On The Structured Manifold Optimization: Reduced-rank and Positive Definite Matrix Estimation

Show full item record

Bookmark or cite this item: http://hdl.handle.net/10027/20247

Files in this item

File Description Format
PDF graph2_1-eps-converted-to.pdf (48KB) (no description provided) PDF
PDF graph2_2-eps-converted-to.pdf (46KB) (no description provided) PDF
PDF graph2_3-eps-converted-to.pdf (48KB) (no description provided) PDF
PDF graph2_4-eps-converted-to.pdf (45KB) (no description provided) PDF
PDF Yuan_Ting.pdf (750KB) (no description provided) PDF
Title: On The Structured Manifold Optimization: Reduced-rank and Positive Definite Matrix Estimation
Author(s): Yuan, Ting
Advisor(s): Hedayat, Samad
Contributor(s): Wang, Junhui; Yang, Jie; Wang, Jing; Martin, Ryan
Department / Program: Mathematics, Statistics, and Computer Science
Graduate Major: Mathematics
Degree Granting Institution: University of Illinois at Chicago
Degree: PhD, Doctor of Philosophy
Genre: Doctoral
Subject(s): Structured matrix Statistical Machine Learning Reduced-rank Model Multi-label Classification Coordinate Descent Positive Definite Matrix Estimation
Abstract: This thesis mainly proposes optimization schemes regarding both types of structured matrices, as well as their applications in statistics. The first structured matrix optimization is proposed under the reduced-rank constraint, and particularly it can be applied for conducting multi-label classification and variable selection simultaneously. The proposed algorithm for optimization is computationally efficient and delivers superior numerical performance in terms of both classification and variable selection accuracy. The asymptotic consistencies are also established to support the advantages of the proposed method. The second structured matrix optimization proposes the estimation of sparse positive definite matrix generated by a generic coordinate descent (CD) algorithm, and particularly it can be applied to the estimation of the high-dimensional covariance matrix and inverse covariance or precision matrix with variable selection.
Issue Date: 2016-02-17
Genre: thesis
URI: http://hdl.handle.net/10027/20247
Rights Information: Copyright 2015 Ting Yuan
Date Available in INDIGO: 2018-02-18
Date Deposited: 2015-12
 

This item appears in the following Collection(s)

Show full item record

Statistics

Country Code Views
United States of America 76
China 71
Ukraine 14
Germany 8
Russian Federation 6

Browse

My Account

Information

Access Key