Determination breast cancer accuracy using data mining

Breast cancer disease is the world's second leading killers among cancer causing death in women. Regardless of the fact that cancer is curable and preventable in early stages, still there are patients who have been diagnosed in later stages. The proposing paper corroborates several detecting and diagnosing methods of cancer, although fully depended on medical technicians and with medical image supporting technique one can detect the cancer causing symptoms in all the stages specifically on later stages. The work's objective is to establish the features and possibilities to achieve accuracies in the breast cancer as either benign or malignant. The work, explores decision trees applicability in predicting the occurrence of breast cancer. Finally performance evaluation analysis is made on several conventional learning algorithms viz. SMO (Sequential Minimal Optimization), Random Forests, J48, Random tree and Naive Bayes. The Investigations and Experiments proved that SMO placed top most with higher accuracy. Based on Experimental evaluations, SMO classifier enhances accuracy 2%, precision 0.018, recall 0.014 and F-measure 0.08 of the proposed classifier compared than previous classifiers.