Open Access Open Access  Restricted Access Subscription or Fee Access

Reduced Support Vector Machine One against One for Multiclass Large Data

Yuniati Dian Pertiwi, Santi Wulan Purnami

Abstract


Support Vector Machine (SVM) is novel technique of machine learning and having good performance in classification problem. However, there is debility for classification large data is inefficient. Various researches that related to large data has also been done, such as Reduced Support Vector Machine (RSVM). But, it is limited for binary classification problem. The aim of this paper is how to contend a large data and multi class problem. In this paper, Reduced Support Vector Machine One-Against-One (RSVM-OAO) was purposed to solve large data and multi class problem. The result of RSVM-AOA method would be compared with Smooth Support Vector Machine One-Against-One ( SSVM-OAO) and Random Sampling SSVM-OAO. Evaluation was done on the classification accuracy and computation time. The result of this study, RSVM-OAO method was resulting 90% classification accuracy and shortly computation time. In SSVM-OAO method, not only resulting classification accuracy more than 90%, but also long computation time. Whereas, the Random Sampling SSVM-OAO method resulting lower classification accuracy and shortly computation time. If three of methods were compared, then RSVM-OAO is better than other, because it resulting high accuracy and short computation time.

Keywords


RSVM, Multiclass, Classification, OAO.

Full Text:

PDF


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information. Also: DOI is paid service which provided by a third party. We never mentioned that we go for this for our any journal. However, journal have no objection if author go directly for this paid DOI service.