Swarm-intelligent elephant herding optimized support vector machine for predicting heart disease

Abstract

Considering the significant worldwide mortality rate of cardiac disease, early and precise prognosis is essential to enhancing patient outcomes. The accuracy, effectiveness, and dependability of the heart disease prediction techniques currently in use, however, continue to encounter difficulties. Therefore, it is essential to create novel strategies that can get beyond these constraints and offer a better way to forecast cardiac disease. A novel advanced elephant herding optimised support vector machine (AEHO-SVM) technique utilized for the study in suggest method for predicting cardiovascular disease. A heart disease dataset is used to assess the efficacy of the suggested AEHO-SVM algorithm. To deal with missing values, normalise the data, and minimise dimensionality, the dataset is pre-processed. The AEHO-SVM approach aims to maximise classification accuracy while lowering the possibility of misdiagnosis by optimising the hyperplane separation and changing the support vectors iteratively. A number of metrics, like sensitivity, accuracy, f-measure, and specificity are used for assess model's performance. The experiment show that the AEHO-SVM technique works better than conventional methods and achieves good predictive accuracy for the diagnosis of heart disease. A suggested AEHO-SVM technique makes use of the collective thinking of elephant herding and its built-in optimisation skills to provide a more effective and trustworthy prediction method.