A new improved FSVM algorithm based on SVDD. (19th November 2018)

• Record Type: Journal Article
• Authors: Xie, Wenhao; Liang, Gongqian; Guo, Qiao
• Other Names: Mei Lin guestEditor.; Yan Zhiguo guestEditor.; Shao Jie guestEditor.
• Is Part Of: Concurrency and computation. Volume 31:Number 9(2019)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2018-11-19
• Subjects: classification hyperplane -- De‐noising algorithm -- FSVM -- fuzzy membership -- SVDD -- SVM; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.4893
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: SVM (Support Vector Machine) is a popular machine‐based learning method, with wide applicability and excellent generalization performance in a large number of real‐time classification problems. However, SVM is sensitive to the noise and outliers. As an improved algorithm based on SVM, FSVM (Fuzzy Support Vector Machine) gives the training samples different fuzzy membership values in order to reduce the interference of the noise and outliers. However, like the normal SVM algorithm, the FSVM algorithm still needs to solve the problems such as how to improve the accuracy of classification and how to accurately recognize the noise and outliers. In this paper, an improved FSVM of data classification algorithm (IFSVM) has been proposed. Firstly, this algorithm deletes the outliers or noise based on the average density algorithm and removes them from the samples, thus avoiding the influence of the noise to the classification accuracy. Secondly, the centers and radiuses of the two minimum hyperspheres are extracted based on the SVDD algorithm. Finally, this algorithm sets the membership function values by comparing the distance between each sample and the center of the sample's hypersphere, the distance between the sample and the opposite hypersphere center, and the distance between the two hypersphere centers. In this way, this algorithm highlights the importance of the boundary vectors, which could be support vectors for classification, and improves the classification accuracy. The experimental results show that this algorithm has improved the anti‐noise ability and the accuracy of classification when compared with the normal SVM algorithm and three other FSVM algorithms.