Training algorithms for fuzzy support vector machines with noisy data
Resource
Pattern Recognition Letters 25 (14): 1647-1656
Journal
Pattern Recognition Letters
Journal Volume
25
Journal Issue
14
Pages
1647-1656
Date Issued
2004-01
Author(s)
Chun-Fu Lin
Abstract
The previous study of fuzzy support vector machines (FSVMs) provides a method to classify data with noises or outliers by manually associating each data point with a fuzzy membership that can reflect their relative degrees as meaningful data. In this paper, we introduce two factors in training data points, the confident factor and the trashy factor, and automatically generate fuzzy memberships of training data points from a heuristic strategy by using these two factors and a mapping function. We investigate and compare two strategies in the experiments and the results show that the generalization error of FSVMs is comparable to other methods on benchmark datasets. The proposed approach for automatic setting of fuzzy memberships makes the FSVMs more applicable in reducing the effects of noises or outliers. © 2004 Elsevier Ltd. All rights reserved.
Other Subjects
Algorithms; Conformal mapping; Database systems; Error analysis; Heuristic methods; Learning systems; Optimization; Pattern recognition; Problem solving; Datasets; Fuzzy membership; Optimization and classification; Support vector machines (SVM); Fuzzy sets
Type
journal article
File(s)![Thumbnail Image]()
Loading...
Name
34.pdf
Size
304.05 KB
Format
Adobe PDF
Checksum
(MD5):08ffa188e60596ef47513a678f45c90c