DEVELOPMENT OF A CARDIOVASCULAR HEART DISEASE CLASSIFICATION MODEL USING NEURAL NETWORK BASED DATA MINING TECHNIQUE

Subject Area: COMPUTER SCIENCE

This paper presents the development of a cardiovascular heart disease classification model using neural network based data mining technique. The aim was to develop a classification model for the detection of cardiovascular heart disease using data mining technique. The research methods are data collection, data extraction, feed forward neural network, and the cardiovascular classification model. The study collected data of 18 heart disease attributes with updated features such as alcohol and kola-nut which is domicile in the African region and has contributed to cardiovascular heart problem in the region. The data was extracted using Best First Search Method and then trained with neural network model. The models were implemented with neural network toolbox in Simulink and evaluated. The result achieved a classification accuracy of 96.51%; Mean square error value of 0.0356810Mu and True positive classification value of 0.959. The result was compared with existing state of the model and a percentage improvement of 2.26% was achieved.