A PROBABILISTIC BAYESIAN CLASSIFIER APPROACH FOR BREAST CANCER DIAGNOSIS AND PROGNOSIS
Sadeghi Hesar Alireza · 2012
Basically, m edical dia gnosis problems ar e the most eff ective com ponent of treatment policies. Recently, signif icant a dvances have been formed in medical diagnosis fields using data mining techniques. Data mining or Knowledge Discovery is sear ching large da tabases to discov er patterns a nd ev aluate the probability of next occurr ences. In this paper, B ayesian Cla ssifier is used as a Non-linear data mining tool to determine the seriousness of breast cancer. The recorded observations of the Fine Needle A spiration (FNA ) tests that ar e obta ined at the University of W isconsin are consider ed a s ex perimental da ta set in this research. The Tabu search algorithm for structural learning of bayesian classifier and Genie simulator for parametric lear ning of bayesian cla ssifier were used. Finally, the obtained results by the proposed model were com pared w ith actual r esults. The Comparison process indicates that seriousness of the disease in 86.18% of cases are guessed very close to the actual values by proposed model.