Constructing L2-SVM-Based Fuzzy Classifiers in High-Dimensional Space With Automatic Model Selection and Fuzzy Rule Ranking

Abstract

In this paper, a new scheme for constructing parsimonious fuzzy classifiers is proposed based on the L2support vector machine (L2-SVM) technique with model selection and feature ranking performed simultaneously in an integrated manner, in which fuzzy rules are optimally generated from data by L2-SVM learning. In order to identify the most influential fuzzy rules induced from the SVM learning, two novel indices for fuzzy rule ranking are proposed and named as αvalues and ω -values of fuzzy rules in this paper. The α -values are defined as the Lagrangian multipliers of the L2-SVM and adopted to evaluate the output contribution of fuzzy rules, while the ω -values are developed by considering both the rule base structure and the output contribution of fuzzy rules. As a prototype based classifier, the L2-SVM based fuzzy classifier evades the curse of dimensionality in high-dimensional space in the sense that the number of support vectors, which equals the number of induced fuzzy rules, is not related to the dimensionality. Experimental results on high-dimensional benchmark problems have shown that by using the proposed scheme the most influential fuzzy rules can be effectively induced and selected, and at the same time feature ranking results can also be obtained to construct parsimonious fuzzy classifiers with better generalization performance than the well-known algorithms in literature.

Type-1 OWA operator based non-stationary fuzzy decision support systems for breast cancer treatments

“…In our study of designing a non-stationary fuzzy expert system for breast cancer treatments, 12 initial fuzzy rules are acquired [11] according to the professional clinical guidelines provided by Nottingham University Hospitals NHS Trust Breast Directorate, i.e., the fuzzy rule base is obtained from human experts' knowledge, which is different from the scheme of inducing fuzzy rules from a dataset [9] [10]. These guidelines include various treatment decisions based on many patients' assessment results (corresponding to the attributes described in the Section II).…”

Automated Group Decision Support Systems Under Uncertainty: Trends and Future Research

“…Liang and Mendel originally suggested a method of designing parsimonious type-2 fuzzy model by using the SVD-QR with column pivoting algorithm to perform rule reduction [60]. However, some research on type-1 fuzzy models [111] indicates that the rule reduction by the SVD-QR with column pivoting algorithm heavily depends on the estimation of an effective rank. The problem is that different estimates of the rank often produce dramatically different rule reduction results.…”

Evolutionary Instance Selection Algorithm based on Takagi-Sugeno Fuzzy Model

“…Moreover, the large amount of data that is available in any research field poses new problems for data mining and knowledge discovery methods. Data reduction is a data preprocessing task that can be applied to ease the problem of dealing with large amounts of data [10]. The best known data reduction processes are feature selection and instance selection.…”