A Liver Disease Diagnosis System Using Data Mining and Fuzzy Logic Techniques

• Main Authors: Chung-Yi Chou, 周忠毅
• Other Authors: Jenn-Long Liu
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/85773918592989938467

碩士 === 義守大學 === 資訊管理學系 === 102 === This study uses classification decision tree algorithm to apply to the prediction of liver disease. By way of the analysis of the cases of database, we can identify the significant attributes which have the effects on causing the liver disease. This study adopts a total of 583 sample data accessed from ILPD (Indian Liver Patient Dataset) in UCI repository database as the disease database. In this study, we apply attribute analysis to find out the significant attributes. By retaining the attributes with high significance to the disease and deleting those with low significance to the disease, the accuracy of prediction can be improved. Moreover, this study uses fuzzy theory to build up a fuzzy expert system, and then we combine the fuzzy expert system and the database of liver disease to create fuzzy rules by using the key attributes of liver disease. Thereafter, a user can enter the inputs of corresponding physiological indices into the system to create the appropriate fuzzy rules and obtain the risk assessment of liver disease. The results of this study not only can help the public to do the prediction by themselves in order to sure whether they are affected with liver disease or not, but also the study can provide a physician for a reference when he/she makes a medical diagnoses to a patient with liver disease.