Classification of Heart Disease Hungarian Data Using Entropy, Knnga Based Classifier and Optimizer

Abstract

To mine the useful information from massive medical databases data mining plays as imperative role. In data mining classification (supervised learning) which can be used to design model by describing significant data classed, where class attribute is involved in the construction of the classifier. In this work, we propose a methodology in which uses KNN classifier. It is simple, popular, more efficient and proficient algorithm for pattern recognition. The samples of the medical databases are classified on the basis of nearest neighbor in which medical database are massively found in nature and contains irrelevant and redundant attributes. The only KNN classifier produce less accurate results that is why we use hybrid approach of KNN and genetic algorithm (GA) to obtain more accurate results. To evaluate the performance of the proposed approach Hungarian dataset (UCI learning) is used to classify the attributes of heart disease. The genetic algorithm performs global research on complex large and multimodal landscapes which provide minimal solutions or search space. The experimental outcomes of accuracy parameter of proposed approach give more accurate and efficient results than the existing approach.