A Study of Applying Data Mining Technology on Hereditary Genes of Colon Cancer

• Main Authors: YI-HUA, TSAI, 蔡依樺
• Other Authors: PANG-CHANG HSIEH
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/61654591471026428184

碩士 === 輔仁大學 === 統計資訊學系應用統計碩士班 === 101 === The colon-cancer cases increase rapidly in Taiwan with 9,000 additional annually. Generally think the colon cancer patients rapidly increasing due to the westernization of lifestyles. But part of the colon cancer patients are caused genetically, that is due to the DNA repair gene mutation. In this study, colon cancer samples from NCBI microarray database were used to investigate the differential expression of genes. This microarray data and corresponding lifestyle information was compared with data of colon cancer echoes from National Health Insurance Database. The resulted decision trees and logistic regression model were constructed with various data mining algorithm. Those genes significantly associated with colon cancer were found out simultaneously. The results showed that the discriminant model is the best method to predict the colon cancer patients. In conclusion, we hope that the Standard Operation Procedure can provide the reference of medical research.