文章基本信息

标题：THE CLASSIFICATION PERFORMANCE USING LOGISTIC REGRESSION AND SUPPORT VECTOR MACHINE (SVM)
本地全文：下载
作者：AGUS WIDODO ; SAMINGUN HANDOYO
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：19
页码：5184
出版社：Journal of Theoretical and Applied
摘要：In the global world, data processing will have a key role for an organization in winning a competition because it will produce the useful information. The mathematical modeling in practice must be able to answer the challenging of information needed by users such as object classification. Many researchers from the various field of study have implementation and development the methods of classification in the real world. The popular classification methods are logistic regression and Support Vector Machine (SVM). This paper will investigate comparison in performance of both methods fairly using to actions, three types background of the data set and transformation to categorial scale for all predictor variables. The performance of both methods will be evaluated using Apparent Error Rate (Aper) and Press�Q statistic. Before modeling process, we divided each data set to become training data that have 80% part of data set and the remain as testing data. In this paper, we successfully show that the SVM has the performance of classification better than logistic regression not only in both training and testing data but also in three difference types and background of data set.
关键词：Aper; classification; logistic Regression; SVM