Title: Bayesian estimation of mean and variance models with penalized splines
Authors: Hector Zarate - Banco de la Republica de Colombia (Colombia) [presenting]
Abstract: The fusion among various statistical methods is extended to estimate the mean and variance functions in semiparametric models when the response variable comes from an exponential family distribution. We rely on the natural connection among penalized regression splines that uses basis functions with generalized linear models and Bayesian Markov Chain sampling simulation methodology. The significance and implications of our strategy lie in its potential to contribute to a simple and unified computational methodology that will take into account the factors that affect the variability of the responses which in turn is important for efficient estimation and correct inference for mean parameters without the requirement of fully parametric models. A simulation study investigates the performance of the estimates. Finally, an application to the LIDAR data highlights the merits of our approach.