Predicting breast cancer recurrence using effective classification and feature selection technique

Abstract

Breast cancer is a major threat for middle aged women throughout the world and currently this is the second most threatening cause of cancer death in women. But early detection and prevention can significantly reduce the chances of death. An important fact regarding breast cancer prognosis is to optimize the probability of cancer recurrence. This paper aims at finding breast cancer recurrence probability using different data mining techniques. We also provide a noble approach in order to improve the accuracy of those models. Cancer patient's data were collected from Wisconsin dataset of UCI machine learning Repository. This dataset contained total 35 attributes in which we applied Naive Bayes, C4.5 Decision Tree and Support Vector Machine (SVM) classification algorithms and calculated their prediction accuracy. An efficient feature selection algorithm helped us to improve the accuracy of each model by reducing some lower ranked attributes. Not only the contributions of these attributes are very less, but their addition also misguides the classification algorithms. After a careful selection of upper ranked attributes we found a much improved accuracy rate for all three algorithms.