ABSTRACT
Cases of cancer in various forms, especially breast cancer among women, have been on the rise in recent times. If we are able to correctly categorize patients having malignant or benign form of breast cancer on the basis of historical data, it would not only eliminate human errors that often get introduced in proper identification of diseases but would also go a long way in providing correct diagnosis and ensure speedy recovery. Although there are a number of classification algorithms available to carry out this task on a given dataset, this chapter focuses on support vector machine (SVM) classifier and K-nearest neighbor classifier. Both techniques are applied on given data and results are recorded. This is followed by a comparative analysis of both the algorithms in achieving the task at hand using appropriate parameters. The chapter concludes with an overview of suggestions on how to obtain better classification results and the scope of reliance on data to arrive at concrete results for the diagnosis of patients suffering from critical ailments. The Breast Cancer Wisconsin (Diagnostic) Data Set is used for experimentation purposes.
