- 期刊
- OpenAccess
應用與比較多種預測技術於光學膜產業銷售額預測
Comparing Different Prediction Techniques for Sales Amount Forecasting in Optical Film Industry
摘要
隨著智慧型通訊裝置的蓬勃發展，屬於相關產品的光學膜（Optical film）市場也隨之成長，並且越來越重要，目前已有許多的公司都進入了光學膜市場。銷售額預測對企業在營運上是一個重要的議題。因為銷售額代表了公司的市場佔有率，若能夠做出正確的銷售額預測，有助於提升企業的營運效率與評估企業的經營績效和營運前景。為了找出適用於光學膜產業銷售額的預測方法，本研究應用天真預測法（Naïve forecast）、逐步迴歸法（Stepwise regression, SR）、倒傳遞類神經網路（Back-propagation neural network, BPN）、支援向量迴歸（Support vector regression, SVR）及極限學習機（Extreme learning machine, ELM）等五種分屬統計與機器學習技術的不同預測方法對光學膜產業的銷售額進行預測。本研究以三家股票上市公司的光學膜相關公司－長興化工、華立企業與瑞儀光電為研究對象，並以單期預測（one-step ahead forecast）及多期預測（multi-step ahead forecast）為預測目標進行分析。實驗結果顯示，SVR在三個公司的資料中，不管是針對單期預測或是多期預測都能有最佳的預測績效，因此本研究建議可使用SVR建構光學膜銷售額預測模式。
並列摘要
Optical film is one of the key components of liquid crystal displays. As liquid crystal displays are widely used in a variety of information processing devices (eg, smart phones, tablet PCs, notebook computers, navigation devices, etc.), optical film industry is an attractive sector to invest in along with the increasing demand of information processing devices. Sales is the most critical function in any company. Therefore, in order to improve the quality of investment strategy for the companies in optical film industry, how to construct an accuracy and reliable sales forecasting model is an important issue. In this study, five different forecasting methods including naive forecast (NF), stepwise regression (SR), support vector regression (SVR), extreme learning machine (ELM) and back propagation neural network (BPN) are used to forecast sales amount of optical film industry. The monthly sales amount data collected from three issued optical film companies in Taiwan are used as experimental data to evaluate the performance of the five forecasting methods. The experimental results showed that SVR method can provide better forecasting results than that of the NF, SR, ELM and BPN. Thus, SVR is a promising technique for forecasting optical film sales amount.