Low-degree Polynomial Mapping of Data for SVM
碩士 === 國立臺灣大學 === 資訊工程學研究所 === 97 === Non-linear mapping functions have long been used in SVM to transform data into a higher dimensional space, allowing the classifier to separate non-linearly distributed data instances. Kernel tricks are used to avoid the problem of a huge number of features of th...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2009
|
Online Access: | http://ndltd.ncl.edu.tw/handle/85416613440796675619 |
id |
ndltd-TW-097NTU05392033 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-097NTU053920332016-05-04T04:31:31Z http://ndltd.ncl.edu.tw/handle/85416613440796675619 Low-degree Polynomial Mapping of Data for SVM 低階多項式資料映射與支持向量機 Yin-Wen Chang 張瀠文 碩士 國立臺灣大學 資訊工程學研究所 97 Non-linear mapping functions have long been used in SVM to transform data into a higher dimensional space, allowing the classifier to separate non-linearly distributed data instances. Kernel tricks are used to avoid the problem of a huge number of features of the mapped data point. However, the training/testing for large data is often time consuming. Following the recent advances in training large linear SVM (i.e., SVM without using nonlinear kernels), this work proposes a method that strikes a balance between the training/testing speed and the testing accuracy. We apply the fast training method for linear SVM to the expanded form of data under low-degree polynomial mappings. The method enjoys the fast training/testing, but may achieve testing accuracy close to that of using highly nonlinear kernels. Empirical experiments show that the proposed method is useful for certain large-scale data sets. Chih-Jen Lin 林智仁 2009 學位論文 ; thesis 32 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立臺灣大學 === 資訊工程學研究所 === 97 === Non-linear mapping functions have long been used in SVM to transform data into a higher dimensional space, allowing the classifier to separate non-linearly distributed data instances. Kernel tricks are used to avoid the problem of a huge number of features of the mapped data point. However, the training/testing for large data is often time consuming. Following the recent advances in training large linear SVM (i.e., SVM without using nonlinear kernels), this work proposes a method that strikes a balance between the training/testing speed and the testing accuracy. We apply the fast training method for linear SVM to the expanded form of data under low-degree polynomial mappings. The method enjoys the fast training/testing, but may achieve testing accuracy close to that of using highly nonlinear kernels. Empirical experiments show that the proposed method is useful for certain large-scale data sets.
|
author2 |
Chih-Jen Lin |
author_facet |
Chih-Jen Lin Yin-Wen Chang 張瀠文 |
author |
Yin-Wen Chang 張瀠文 |
spellingShingle |
Yin-Wen Chang 張瀠文 Low-degree Polynomial Mapping of Data for SVM |
author_sort |
Yin-Wen Chang |
title |
Low-degree Polynomial Mapping of Data for SVM |
title_short |
Low-degree Polynomial Mapping of Data for SVM |
title_full |
Low-degree Polynomial Mapping of Data for SVM |
title_fullStr |
Low-degree Polynomial Mapping of Data for SVM |
title_full_unstemmed |
Low-degree Polynomial Mapping of Data for SVM |
title_sort |
low-degree polynomial mapping of data for svm |
publishDate |
2009 |
url |
http://ndltd.ncl.edu.tw/handle/85416613440796675619 |
work_keys_str_mv |
AT yinwenchang lowdegreepolynomialmappingofdataforsvm AT zhāngyíngwén lowdegreepolynomialmappingofdataforsvm AT yinwenchang dījiēduōxiàngshìzīliàoyìngshèyǔzhīchíxiàngliàngjī AT zhāngyíngwén dījiēduōxiàngshìzīliàoyìngshèyǔzhīchíxiàngliàngjī |
_version_ |
1718259432132968448 |