Abstract:
An accurate forecasting module is a key element of any revenue management system. This module includes demand forecasting, which involves tasks of forecasting complex seasonal time series. The challenge of producing accurate demand forecasts requires the application of suitable forecasting methods to address that complexity. The aim of this paper is to evaluate a new innovation state space modeling framework, based on innovations approach, developed for forecasting time series with complex seasonal patterns. This modeling framework provides an alternative to existing models of exponential smoothing, since it is capable of tackling seasonal complexities such a multiple seasonal periods and high frequency seasonality.
