Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition
碩士 === 國立臺灣大學 === 應用數學科學研究所 === 105 === Dimension reduction and feature extraction are the important techniques in the big-data era to reduce the dimension of data and the computational cost for further data analysis. Low-rank singular value decomposition (low-rank SVD) is the key part of these tech...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2017
|
Online Access: | http://ndltd.ncl.edu.tw/handle/8k8nk9 |
id |
ndltd-TW-105NTU05507008 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-105NTU055070082019-05-15T23:39:40Z http://ndltd.ncl.edu.tw/handle/8k8nk9 Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition 整合多重隨機奇異值分解與理論分析 Da-Wei Chang 張大衛 碩士 國立臺灣大學 應用數學科學研究所 105 Dimension reduction and feature extraction are the important techniques in the big-data era to reduce the dimension of data and the computational cost for further data analysis. Low-rank singular value decomposition (low-rank SVD) is the key part of these techniques. In order to compute low-rank SVD faster, some researchers propose to use randomized subspace sketching algorithm to get an approximation result (rSVD). In this research, we propose an idea for integrating the results from randomized algorithm to get a more accurate approximation, which is called integrated singular value decomposition (iSVD). We analyze iSVD and the integration methods by theoretical analysis and numerical experiment. The integration scheme is a constraint optimization problem with unique local maximizer up to orthogonal transformation. Line search type method, Kolmogorov-Nagumo type average method and reduction type method are introduced and analyzed for their theoretical background and computational complexity. The similarity and difference between iSVD and rSVD with same sketching number are also explained and analyzed. The numerical experiment shows that the line search method in iSVD converges faster than the one in rSVD for our test examples. Also, using the integrated subspace from reduction as the initial value of line search method can reduce the iteration number to converge. 王偉仲 2017 學位論文 ; thesis 54 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立臺灣大學 === 應用數學科學研究所 === 105 === Dimension reduction and feature extraction are the important techniques in the big-data era to reduce the dimension of data and the computational cost for further data analysis. Low-rank singular value decomposition (low-rank SVD) is the key part of these techniques. In order to compute low-rank SVD faster, some researchers propose to use randomized subspace sketching algorithm to get an approximation result (rSVD). In this research, we propose an idea for integrating the results from randomized algorithm to get a more accurate approximation, which is called integrated singular value decomposition (iSVD). We analyze iSVD and the integration methods by theoretical analysis and numerical experiment. The integration scheme is a constraint optimization problem with unique local maximizer up to orthogonal transformation. Line search type method, Kolmogorov-Nagumo type average method and reduction type method are introduced and analyzed for their theoretical background and computational complexity. The similarity and difference between iSVD and rSVD with same sketching number are also explained and analyzed. The numerical experiment shows that the line search method in iSVD converges faster than the one in rSVD for our test examples. Also, using the integrated subspace from reduction as the initial value of line search method can reduce the iteration number to converge.
|
author2 |
王偉仲 |
author_facet |
王偉仲 Da-Wei Chang 張大衛 |
author |
Da-Wei Chang 張大衛 |
spellingShingle |
Da-Wei Chang 張大衛 Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
author_sort |
Da-Wei Chang |
title |
Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
title_short |
Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
title_full |
Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
title_fullStr |
Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
title_full_unstemmed |
Theoretical and Performance Analysis for Integrated Randomized Singular Value Decomposition |
title_sort |
theoretical and performance analysis for integrated randomized singular value decomposition |
publishDate |
2017 |
url |
http://ndltd.ncl.edu.tw/handle/8k8nk9 |
work_keys_str_mv |
AT daweichang theoreticalandperformanceanalysisforintegratedrandomizedsingularvaluedecomposition AT zhāngdàwèi theoreticalandperformanceanalysisforintegratedrandomizedsingularvaluedecomposition AT daweichang zhěnghéduōzhòngsuíjīqíyìzhífēnjiěyǔlǐlùnfēnxī AT zhāngdàwèi zhěnghéduōzhòngsuíjīqíyìzhífēnjiěyǔlǐlùnfēnxī |
_version_ |
1719152366329004032 |