A Heart Disease Classification using an Improved Particle Swarm Optimization Algorithm with Support Vector Machine Classifier

The Heart Disease (HD) is one of the leading causes of death worldwide. Early detection of HD decreases heart attack disease risk and increases recovery chances. Unbalanced samples and missing values for certain features are the primary challenges in HD classification. This work aims to implemented the effective feature optimization and classification methods for HD classification, which assists in the early diagnosis of the condition by medical professionals. Hence, a HD classification method based on Improved Particle Swarm Optimization (IPSO) algorithm and Support Vector Machine (SVM) categorization method, namely IPSO-SVM method, is implemented to evaluated and predict HD risk. At first, Cleveland dataset’s collected data is rescaled using technique of min-max normalization. Then, ResNet 50 is utilized for extract the features and IPSO is employed to select the significant features. Finally, Machine Learning (ML) classifier used for HD classification, which include both presence and absence of the HD. Compared with existing methods, the implemented method achieved 99.95% of accuracy using Cleveland dataset.