การประยุกต์ใช้ตัวแบบการพยากรณ์โดยเทคนิคโครงข่ายประสาทเทียม: กรณีศึกษาการพยากรณ์ผลผลิตลำไยนอกฤดู

Abstract

Abstract

The problem of a typical seasonal Longan is a difference between demand and supply of it. Thus, off-season Longan is considered as one solution to solve this problem. However, the same problem still occurs due to a lack of information and technique sharing among all members in the Longan supply chain. This study aims to present the application of forecasting model developed based on Full Artificial Neuron Network (FANN) technique for off-season Longan supply forecasting. This model applies 40 input factors that have effects on the outputs of the off-season Longan. The advantage of this model is to help the farmers to forecast their off-season Longan outputs for planning their productions, sales and marketing in the effective way. The model presented in this research work can be improved by updating the input data as present as the time used to maintain the accuracy of the forecasted value comparing with the actual output value.