ARMA Modeling and Forecasting
Fitting an appropriate ARMA(p, q) model to an observed time series data set involves two interrelated problems, namely determining the order (p, q) (which is usually referred to as model identification) and estimating parameters in the model. Further, the postfitting diagnostic checking on the validity of the fitted model is equally important.
KeywordsBayesian Information Criterion Akaike Information Criterion Maximum Likelihood Estimator Standardize Residual ARMA Model
Unable to display preview. Download preview PDF.