A new taxonomy for vector exponential smoothing and its application to seasonal time series

In short-term demand forecasting, it is often difficult to estimate seasonality accurately, owing to short data histories. However, companies usually have multiple products with similar seasonal demand pat-terns. A possible solution in this case is to use the components of several time series from a homoge-neous family, thus estimating seasonal coefficients based on cross-sectional information. Motivated by this practical problem, we propose a new taxonomy of Parameters, Initial States and Components (PIC), which exploits homogeneous features of time series. We then apply this framework to vector exponential smoothing. We develop a model selection mechanism based on information criteria to select the appro-priate PIC restrictions. We then conduct a simulation experiment and empirical analysis on retail data in order to assess the performance of point forecasts and prediction intervals of the models within this framework.(c) 2022 The Author(s). Published by Elsevier B.V.This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

Automated summary

Short-term demand forecasting faces challenges in accurately estimating seasonality due to limited data histories. This study proposes a taxonomy called Parameters, Initial States, and Components (PIC) that leverages the homogenous features of time series. The framework is applied to vector exponential smoothing, and a model selection mechanism is developed to choose the appropriate PIC restrictions.