Foresight: the International Journal of Applied Forecasting (2007), 6, 12-15.
Rob J Hyndman1 and Andrey V. Kostenko2
- Department of Econometrics and Business Statistics, Monash University, Clayton VIC 3800, Australia.
- Independent Complex Systems Consultant, 665714, Bratsk, Russia.
Abstract: How much data do you need to forecast using a seasonal model? The answer depends on the type of model being used and the amount of random variation in the data. We discuss the mathematical limits for estimating various common seasonal forecasting models from data. These limits apply when the amount of random variation is very small. Real data often contain a lot of random variation, and then many more observations are required.
Keywords: time series models, forecasting, seasonality, short data