An Application of GA-SVM in the Prediction of Cotton Price

• Main Authors: Chen-Lung Chen, 陳震隆
• Other Authors: Ming-Hsiang Huang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/78297663648721813564

碩士 === 國立彰化師範大學 === 企業管理學系國際企業經營管理 === 101 === Cotton is an very important materials of textile industry in Taiwan, but it is purely rely heavily from import. The highly volatility of cotton price may results in big losses and further jeopardize the price competiveness of Taiwanese firms. Thus, accurately prediction in the cotton price is an critical issue to the sustainable growth of Taiwanese textile industry. More recently, data mining techniques has become a preferred prediction approach to both academia and industry. The objective of this study is to investigate whether an adoption of a data mining approach, Genetic Algorithms (GA) plus Support Vector Machine (SVM), can improve the accuracy of cotton price prediction. The study utilizes data related to the price volatility of cotton over the period from 2001 to 2012, to formulate the said cotton price prediction model. In the hope to provide useful information to the general public and management of firm in the industry. Empirical finding suggests that the alternative method, GA-SVM outperforms the standard method, BPNN, in terms of prediction accuracy.