Estimation and Performance of Generalized Bilinear Time Series Models in the Presence of 2-1 Subsets

Generalized and subset generalized bilinear models which are capable of achieving stationary for all non-linear series were developed to accommodate error process of elements of 2q-1 possible subsets with a view to achieve better model in the generalized bilinear model. The parameters of the models are estimated using robust nonlinear least-square method and Newton-Raphson iterative method. An algorithm is proposed for fitting error process included (EPI) generalized and subset generalized bilinear models. To determine the order of the models, Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were adopted. The included error process made the generalized bilinear model a better model using the residual variance thereby fitting models for all possible subsets listed to have the best model is not necessary. We illustrated the above concept with a real life data.