Support vector interval regression networks for interval regression analysis

✍ Scribed by Jin-Tsong Jeng; Chen-Chia Chuang; Shun-Feng Su

• Publisher: Elsevier Science
• Year: 2003
• Tongue: English
• Weight: 459 KB
• Volume: 138
• Category: Article
• ISSN: 0165-0114
• DOI: 10.1016/s0165-0114(02)00570-5

✦ Synopsis

In this paper, the support vector interval regression networks (SVIRNs) are proposed for the interval regression analysis. The SVIRNs consist of two radial basis function networks. One network identiÿes the upper side of data interval, and the other network identiÿes the lower side of data intervals. Because the support vector regression (SVR) approach is equivalent to solving a linear constrained quadratic programming problem, the number of hidden nodes and the initial values of adjustable parameters can be easily obtained. Since the selection of a parameter in the SVR approach may seriously a ect the modeling performance, a two-step approach is proposed to properly select the value. After the SVR approach with the selected , an initial structure of SVIRNs can be obtained. Besides, outliers will not signiÿcantly a ect the upper and lower bound interval obtained through the proposed two-step approach. Consequently, a traditional back-propagation (BP) learning algorithm can be used to adjust the initial structure networks of SVIRNs under training data sets without or with outliers. Due to the better initial structure of SVIRNs are obtained by the SVR approach, the convergence rate of SVIRNs is faster than the conventional networks with BP learning algorithms or with robust BP learning algorithms for interval regression analysis. Four examples are provided to show the validity and applicability of the proposed SVIRNs.