ENHANCED APPROACH OF HEART DISEASE DIAGNOSIS USING MODIFIED FEATURE EXTRACTION METHOD

Abstract

Heart diseases are becoming one of the leading causes of death in the arena. Consequently, the medical community has shown a great deal of interest in cardiac disease prediction. In order to assist doctors in the design of clinical procedures, a number of studies have created machine learning algorithms for the early prediction of coronary heart diseases. The function set that is chosen has a significant impact on how well such structures perform. This will be more difficult if the schooling dataset has missing values for the various capacities. It is well known that Principal Component Analysis (PCA) can be used to address the issue of missing attribute data. This study offers a method for identifying heart disease by using scientific testing data as input, identifying coronary heart disease by extracting a low dimensional characteristic subset. The suggested approach uses Modified Principal Component Analysis (M-PCA) to extract improved depth characteristics from fresh projections. PCA reduces the size of the function by assisting in the extraction of projection vectors that significantly contribute to the maximum covariance. Three datasets are analyzed to determine the impact, accuracy, sensitivity, and specificity of the suggested approach. The results obtained from the use of the suggested M-PCA technique are compared to earlier studies in order to demonstrate its relevance. The dataset generated by the suggested M-PCA technique was incredibly accurate.