Adaptive SVM for data stream classification

In this paper, we address the problem of learning an adaptive classifier for the classification of continuous streams of data. We present a solution based on incremental extensions of the Support Vector Machine (SVM) learning paradigm that updates an existing SVM whenever new training data are acqui...

Full description

Saved in:
Bibliographic Details
Published inSouth African computer journal = Suid-Afrikaanse rekenaartydskrif Vol. 29; no. 1; pp. 27 - 42
Main Authors Abdulkarim, Salihu A., Lawal, Isah A.
Format Journal Article
LanguageEnglish
Published Makhanda South African Computer Society (SAICSIT) 01.07.2017
South African Institute of Computer Scientists and Information Technologists
Subjects
Action Classification
Human action classification
Incremental Learning
Spam filtering
Support vector machine
SVM
Online AccessGet full text

Cover

More Information
Summary:In this paper, we address the problem of learning an adaptive classifier for the classification of continuous streams of data. We present a solution based on incremental extensions of the Support Vector Machine (SVM) learning paradigm that updates an existing SVM whenever new training data are acquired. To ensure that the SVM effectiveness is guaranteed while exploiting the newly gathered data, we introduce an on-line model selection approach in the incremental learning process. We evaluated the proposed method on real world applications including on-line spam email filtering and human action classification from videos. Experimental results show the effectiveness and the potential of the proposed approach.
ISSN:1015-7999
2313-7835
DOI:10.18489/sacj.v29i1.414