GIBBS SAMPLING APPROACH TO VARIABLE SELECTION IN LINEAR REGRESSION WITH OUTLIER VALUES

Atilla YARDIMCI, Aydın ERAR
1.618 408

Abstract


ABSTRACT

In this study, Gibbs sampling has been applied to the variable selection in the linear regression model with outlier values. Gibbs sampling has been compared with classical variable selection criteria by using dummy data with different β and priors.


Keywords


Key Words: Bayesian variable selection, prior distribution, Gibbs sampling, Markov Chain Monte Carlo, outlier values, entropy

Full Text:

PDF