Bayesian Variable Selection for Mixture Process Variable Design Experiment

This paper discussed Bayesian variable selection methods for models from split-plot mixture designs using samples Metropolis-Hastings within the Gibbs sampling algorithm. is easy to implement due improvement in computing via MCMC sampling. We described methodology by introducing framework, and explaining Markov Chain Monte Carlo (MCMC) The was used draw dependent full conditional distributions which were explained. In experiments with process variables, response depends not only on proportions of components but also effects variables. many such mixture-process experiments, constraints as time or cost prohibit treatments completely at random. these situations, restrictions randomisation force level combinations one group factors be fixed other are run. Then a new first-factor set computational algorithm Stochastic Search Variable Selection (SSVS) linear mixed models. extended SSVS fit design Split-plot Design (SSVS-SPD). motivation this extension that we have two different levels experimental units, whole plots subplots design.