SAFETY STOCK BASED ON CONSUMPTION FORECAST BY THE ARTIFICIAL NEURAL NETWORK
In this paper, the safety stock based on consumption of components in the jewelry business was investigated using the forecasting capability of an Artificial Neural Network (ANN). Generally, this business also has links with fashion, therefore, rapid change in the fashion industry makes the forecasting situation more complicated. The demand fluctuates with customer requirements and high competition in the marketplace. Factors such as late delivery and bad component quality can cause shortages of components. To prevent shortages, provide support to supply management and enhance customer satisfaction, an ANN is utilized for consumption forecast The experimental results were encouraging with the 0.9737 correlation between actual consumption and predicted consumption from the network. The results of MAD and a tracking signal illustrated that the safety stock calculated from the network forecast are better than the company practice. It is implied that there are significant benefits from the inclusion of the consumption forecasting when determining the optimal safety stocks.
safety stock artificial neural network back propagation tracking signal jewelry business
Naroumon Yordphet Siripun Sanguansintukul
Department of Mathematics Faculty of Science, Chulalongkorn University Bangkok, Thailand
国际会议
2010 International Conference on Software and Computing Technology(2010年软件与计算机技术国际会议 ICSCT 2010)
昆明
英文
254-258
2010-10-17（万方平台首次上网日期，不代表论文的发表时间）