Data Mining for Optimal Combination Demand Forecasts
Chi Kin Chan (The Hong Kong Polytechnic University, Hong Kong), Heung Wong (The Hong Kong Polytechnic University, Hong Kong), Wan Kai Pang (The Hong Kong Polytechnic University, Hong Kong) and Marvin D. Troutt (Kent State University, USA)
Copyright: © 2003
This chapter is a case study in combining forecasts for inventory management in which the need for data mining in combination forecasts is necessary. The need comes from selection of sample items on which forecasting strategy can be made for all items, selection of constituent forecasts to be combined and selection of weighting method for the combination. A leading bank in Hong Kong consumes more than 300 kinds of printed forms for its daily operations. A major problem of its inventory control system for such forms management is to forecast their monthly demand. The bank currently uses simple forecasting methods such as simple moving average and simple exponential smoothing for its inventory demands. In this research, the individual forecasts come from well-established time series models. The weights for combination are estimated with quadratic programming. The combined forecast is found to perform better than any of the individual forecasts. Some insights in data mining for this context are obtained.