NONLINEAR REGRESSION MODELING VIA REGULARIZED GAUSSIAN BASIS FUNCTIONS

概要

論文の詳細を見る
Nonlinear regression modeling based on basis expansions has been widely used to explore data with complex structure. There are various types of basis functions to capture complex nonlinear phenomena. In this paper we introduce nonlinear regression models with Gaussian basis functions, for which new Gaussian bases are constructed, taking advantages of $ B $-spline bases. In order to choose adjusted parameters, we derive model selection and evaluation criteria from information-theoretic and Bayesian viewpoints. Monte Carlo simulations and real data analysis show that our proposed modeling strategy performs well in various situations.