Research Article

Open Access

The Robust Bayesian Approach to the Model Selection Algorithm

Krzysztof W Fornalski*, Ludwik Dobrzyński

Abstract

The paper presents the robust Bayesian approach to the experimental data analysis. Firstly, the algorithm of the curve fitting to data points is introduced. The method is based on the robust Bayesian regression analysis, which substantially reduces the role of outlying data points (outliers). In the second part the Bayesian model selection algorithm was presented. Finally, the exemplary applications of both methods are discussed.