Leave-Two-Out Cross Validation to optimal shape parameter in radial basis functions. (March 2019)

• Record Type: Journal Article
• Authors: Azarboni, Habibe Ramezannezhad; Keyanpour, Mohammad; Yaghouti, Mohammadreza
• Is Part Of: Engineering analysis with boundary elements. Volume 100(2019)
• Page Start: 204
• Page End: 210
• Publication Date: 2019-03
• Subjects: Radial basis functions -- Leave-One-Out Cross Validation -- Approximate Moving Least Squares -- Shape parameter -- Cross validation; Boundary element methods -- Periodicals; Engineering mathematics -- Periodicals; Équations intégrales de frontière, Méthodes des -- Périodiques; Mathématiques de l'ingénieur -- Périodiques; Boundary element methods; Engineering mathematics; Periodicals; 620.00151
• DOI: 10.1016/j.enganabound.2018.06.011
• Languages: English
• ISSNs: 0955-7997

Abstract

Abstract: Determination of shape parameter has a major role in the accuracy of the radial basis functions method. In this paper, we present a new method called Leave-Two-Out Cross Validation to determine the best shape parameter. In the proposed method, by deleting two data from the data set, a new formula is derived for determining the error value and then the optimal shape parameter is determined. Num results show that the method will be more accurate in comparison with other methods.