Bayesian approach to bandwidth selection for multivariate count regression function estimation by associated discrete kernel
Nonparametric regression is an important tool for exploring the unknown relationship between a response variable and a set of explanatory variables also known as regressors. This article introduces the associated discrete kernel for multivariate nonparametric count regression estimation. We propose a Bayesian approach based upon likelihood cross-validation and a Monte Carlo Markov chain (MCMC) method for deriving the global optimal bandwidths. Through simulation and real count data, we point out the performance of binomial and triangular discrete kernels. A comparative study of the Bayesian approach and cross-validation technique is also presented.
KeywordsBayesian approach discrete kernel multivariate kernel regression cross-validation
AMS Subject Classification62G07 62G99
Unable to display preview. Download preview PDF.