A MODIFIED BOOTSTRAP METHOD FOR INTERMITTENT DEMAND FORECASTING FOR RARE SPARE PARTS

作者：Jung Gisun; Park Jinsoo; Kim Yohan; Kim Yun Bae^*

来源：International Journal of Industrial Engineering-Theory Applications and Practice, 2017, 24(2): 245-254.

• intermittent demand
• non-parametric forecasting
• bootstrap
• inventory management

An effective inventory management requires timely and accurate forecasting of demand for parts or items. In many real-world scenarios, however, the demand for rare, high-cost spare parts are scarce and erratic, making it highly challenging to perform a reliable forecast for intermittent demand. Studies in the past offered two approaches to such intermittent demand forecasting: a traditional approach to estimating demand parametrically, and a non-parametric approach that estimates the distribution of demands. The bootstrap method is considered to be one of the key non-parametric methods available. Despite its usefulness, however, application of conventional bootstrap methods in intermittent demand forecasting does not take into account any existing dependent structures in lead time demand, leading to inaccurate forecasting. In this paper, we suggest a new bootstrap method that takes into consideration the unique characteristics of intermittent demand to improve forecasting performance. We conclude by demonstrating the applicability of suggested new method through the results of a simple case experiment.

• 出版日期2017