Bayesian adaptive control in linear regression model with informative prior

Abstract

In the Bayesian approach, the prior information about parameters or models can be flexibly and formally incorporated in the analysis of control problems which involve optimisation over several time periods with stochastic elements when uncertainty about parameters is present. We have obtained the optimal value for the independent variable in the immediate first future period using informative priors. A comparison of our results has been done with chose obtained by Zellner (1971) where he has made use of diffuse priors.