New theory revolutionizes time series model estimation for accurate predictions.
The article presents a new theory for estimating time series models. The researchers show that certain types of estimators are reliable and accurate for predicting future values in these models. They tested this theory on various models, like the long-memory fractional ARIMA model and the GARCH(1,1) model, and found that the estimators worked well in these cases. This research helps us better understand and predict patterns in time series data.