Data Mining in the Application of Medical Industry-Empirical Results of Forecasting the Influenza-Like Patients in Taipei

• Main Authors: Kuei-Yung Chang, 張桂源
• Other Authors: Tian-Shyug Lee
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/12307106430783859279

碩士 === 輔仁大學 === 管理學研究所 === 93 === In 2003 the outbreak of the Severe Acute Respiratory Syndrome (SARS) results a severe threat to the whole world. The government and compatriots have begun to pay more attention to the influenza-like illness occurred every year in Taiwan. When establishing the information systems and policies of epidemic prevention, it’s important to accurately determine the epidemic trend first. Data mining techniques are very popular and have been widely applied in different research areas these days. The objective of this study is to build a forecasting model of the numbers of influenza-like illness by integrating multivariate adaptive regression splines (MARS) with artificial neural networks (ANNs) and support vector machine (SVM). The rationale of the study is firstly to build a MARS prediction model, the obtained significant variables of MARS prediction models are then served as the input variables of the ANNs model and SVM model. In order to verify the feasibility and effectiveness of the proposed approaches, the weekly experimental data of the patients of influenza-like illness from 1999 to 2004 was used in this study. As the results revealed, the proposed two hybrid models have better forecasting results and hence provide efficient alternatives in building the influenza-like patients forecasting models.