Medical Data Mining Using Efficient Qpso-Fcm Clustering & Hybrid Svm-Decision Tree Classification Techniques

Abstract

Conventional medical or health care services are rapidly shifting to the internet with the rise of Internet of Health (IoH) era and have been generating a significant measure of health data related to medicine, medical infrastructure, doctors, patients, and so on. The health care services benefit from the effective analyses of these IoH results. Data mining and information discovery is a recent, fundamental research area which has significant applications in medicine, education, science, engineering and industry. It is a method of calculating and determining useful information from a large data set. The goal of data mining is to create, analyze, and apply simple induction processes that make it easier to extract useful knowledge and information from unstructured data. An effective clustering technique aids in partitioning a dataset into many groups, with the similarity in each group being higher than the similarity between groups. In this paper, the Fuzzy C-Means Clustering algorithm is combined with the Quantum-behaved Particle Swarm Optimization (QPSO). The QPSO algorithm's global search capacity helps to prevent local optima stagnation, whereas FCM's soft clustering method helps to divide the data on the basis of membership probabilities. Data classification is a crucial technique for extracting useful data. In this paper, a hybrid classification method is proposed that aims to combine the benefits of both decision trees and support vector machine (SVM) to produce better classification results. The proposed approach reduces the training dataset for SVM classification by using decision tree algorithm and it produces faster results with higher accuracy rates.