Forecasting Monthly Sales Volume of the Farmers' Market by Developing a Seasonal Adjustable Simulator

概要

論文の詳細を見る
The sales analysis and forecast are intended to be useful for farmers and farm direct marketing business. The purpose of this research is to provide efficient, effective, and inexpensive information technology solutions and services to improve the competitiveness of farmers' markets; therefore, the computation speed and the calculation of reliable routine forecasts are important tasks. The main objective of this research is to develop a prototype forecasting simulator which is able to tolerate the information incomplete circumstances. In addition, this prototype simulator has a feature that requires no human intervention to initiate and process model optimization procedures. In general, the exponential smoothing method provides a high level of accuracy for short-range forecasts. In addition, it has minimal data requirements and can be implemented fully automatic. Therefore, it is a good method to use when forecasts are required for large numbers of items. Empirical results were encouraging, with seasonal adjustments the accuracy improved significantly.