The topic of the previous lecture is continued. A sequential approach to the estimation of the underlying parameter will be presented. First results on pseudo-maximum-likelihood-estimators \hat{theta}_T based on discrete time observation are presented.