PREDICTION OF SURVIVAL IN PATIENTS WITH BREAST CANCER USING THREE ARTIFICIAL INTELLIGENCE TECHNIQUES
Chengtao Yu, Cheng‐Min Chao, Bor‐Wen Cheng · 2014
As medical technology advances, has accumulated a large number of health-related data. Faced with incr easingly complex analytical requirements, predictive data mi ning has become an essential instrument for hospita l management and medical research. In this study, the breast can cer dataset is collected from a regional teaching h ospital in central Taiwan between 2002 and 2009. The prognostic factors composed of 8 attributes including 967 subjects, of which 861 are survival after treatment. The three techniques, artificial neural networks (ANNs), support vector machine (SVM) and Bayesian classifier, have been discussed which is used to investigated and evaluated for predictin g breast cancer survival. As can be seen from the results, the pred iction accuracy of a 10-fold cross validation is 90 .31%, 89.79% and 88.64%, respectively. Classification results of SVM are slightly better as compared to ANN and Bayesia n classifier, however, from a relatively low variance, the result s show that the SVM will be the best prognosis in c linical practice.