Academic Journals Database
Disseminating quality controlled scientific knowledge

Improved Double Memberships of Fuzzy Support Vector Machine

ADD TO MY LIST
 
Author(s): WU Xiao | WEI Yan | WU Xia

Journal: Journal of Chongqing Normal University
ISSN 1672-6693

Volume: 28;
Issue: 5;
Start page: 49;
Date: 2011;
Original page

Keywords: SVM | double memberships | C-mean clustering

ABSTRACT
In traditional support vector machine ,there is an existence of noise and outlier sensitive, and often has an over-fitting problem. In this paper, a new membership of fuzzy support vector machine desigh method is proposed-double memberships of fuzzy support vector machine(DM-FSVM),Not only consider the distance of samples to center, but also divide the sample points into two types according to the distance of sample points to the center, near the clustering center sample's membership determined by the distance from this sample to the clustering center, but for the other samples which are far away from clustering center, its membership is a ratio of samiler and disparate class points within its neighborhood. Meanwhile fuzzy vector machine widespread exists long training time probleml. using cut sets C-mean clustering method for training data to clustering process and treat the clustering center as new samples to training, Experimental results show the good performance of the present approach.
Why do you need a reservation system?      Save time & money - Smart Internet Solutions