Bayesian linear discriminant analysis for breast cancer classification
Harikumar Rajaguru, Sunil Kumar Prabhakar · 2017
One of the prominent cancers affecting more women in this world is breast cancer. When a particular cell or specific group of cells in breast grow in an uncontrolled manner and tries to spread by multiplying, then it causes breast cancer. Such an activity often results in tumour, neoplasm or mass which may be benign or malignant. In benign cases, the abnormal growth of cells is restricted to a single group of cells. In malign cases, the abnormal growth of cells invades the tissues surrounding them and even spreads to the remote areas of the body. Artificial intelligence techniques play a vital role in contributing a lot in the advancement of treatment of breast cancer. With the advent of machine learning and soft computing techniques, a lot of reliable and robust technologies have been developed to predict and classify the risk of breast cancer. In this work, Bayesian Linear Discriminant Analysis Classifier (BLDA) is used to classify the risk of breast cancer in the patients and the results are shown in terms of classification accuracy, performance index, sensitivity and specificity. The results show that an average classification accuracy of about 83.45% is reported when Bayesian Linear Discriminant Classifier is utilized.