Nonparametric Bayesian approach to the detection of change point in statistical process control Issah

Abstract

This paper gives an intensive overview of nonparametric Bayesianmodel relevant to the determination of change point in a process control.We rst introduce statistical process control and develop on it describingBayesian parametric methods followed by the nonparametricBayesian modeling based on Dirichlet process. This research proposes anew nonparametric Bayesian change point detection approach which incontrast to the Markov approach of Chib [6] uses the Dirichlet processprior to allow an integrative transition of probability from the posteriordistribution. Although the Bayesian nonparametric technique on themixture does not serve as an automated tool for the selection of thenumber of components in the nite mixture. The Bayesian nonparametricmixture shows a misspecication model properly which has beenexplained further in the methodology. This research shows the principalstep-bystep algorithm using nonparametric Bayesian technique withthe Dirichlet process prior dened on the distribution to the detectionof change point. This approach can be further extended in the multivariatechange point detection which will be studied in the near future.