Classification algorithms in Data Mining

Abstract

Data mining is a relevant term that simplifies the exploration and analysis of the huge amount of data with the aim of looking for hidden and valuable information from it. The area of interest among researchers in involving Data mining approaches to handle healthcare datasets have increased recently. In this context, we tried to categorize patients; who can be affected by breast cancer, diabetes or hypothyroidism; according to the stage of each disease. In fact, several classification algorithms; including SimpleLogistic, Instance-based k-nearest Neighbors (IBK), Naive Bayes, Stochastic Gradient Descent (SGD), Logistic model tree (LMT) and Sequential Minimal Optimization (SMO), are compared in terms of powerful performance measures. Therefore, after using the well-known data mining and knowledge discovery tool; Weka, we managed to pick the classification technique; LMT, which proved its high accuracy and scalability when dealing with healthcare datasets having different characteristics.