I am performing a univariate and multivariate time-series analysis. After identifying the best ARIMA models for each individual time-series, I performed the cross-correlation function to identify the lags to be included in the dynamic regression model. After performing the transfer function model (TFM) with the identified lags, there is no autocorrelation of the residuals. I was wondering if, despite that, it is necessary to include the AR and MA parameters in the TFM identified during the univariate analysis?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.