A Dependent-Multivariate Data Analysis Method by Skewed-RBF Network Based on FGM Copula  
Author Shuhei Ota

 
Co-Author(s) Mitsuhiro Kimura

 
Abstract We extend the traditional RBF (radial basis function) network [1] to be more powerful in terms of treating the dependent input variables. For this purpose, the idea of copula [2] function is introduced. After proposing the new model, we compare the estimation performances among three models (new model, traditional RBF network, and multiple regression model) via the sample data analysis. We show the advantage of our new model by the numerical experiments.

 
Keywords FGM copula, RBF network, multiple regression analysis
   
    Article #:  21256
 
Proceedings of the 21st ISSAT International Conference on Reliability and Quality in Design
August 6-8, 2015 - Philadelphia, Pennsylvia, U.S.A.