Hybrid intelligent system for demand forecasting in die-casting industry

• Main Author: Makhtar, Nooraini
• Format: Thesis
• Language: English
• Published: 2015
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://eprints.utm.my/id/eprint/48716/25/NoorainiMakhtarMFKM2015.pdf; http://eprints.utm.my/id/eprint/48716/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:85355

Forecasting is one of the important elements in business nowadays. An accurate forecast of future demand is an absolute requirement for planning production without creating wasteful overages or shortages. The accurate forecast is very importance for industry especially Electronics Devices Industry. As many knows, Electronic Devices Industry is promised the fluctuate demand. To compete the forecast with the demand, many industry had to choose the hybrid Intelligent System forecast model rather than stand alone model. For this case study, long memory forecast model is hybrid with the Artificial Intelligent model are chosen to forecast the demand one of the electronic devices supplier company. Auto Regression Fractional Integrated Moving Average (ARFIMA) are chosen as the long memory process model and hybrid with Artificial Neural Network (ANN) model as the Artificial Intelligent. The hybrid Intelligent System Model improve the forecast error for the next year of demand by using the current demand with 1.4% forecast error.