Earlier identification of heart disease using enhanced genetic algorithm and fuzzy weight based support vector machine algorithm

Abstract

The diagnosis of heart disease is a huge challenge capable of providing an automated prediction on the stage of heart disease with the intent of proactively taking that countermeasure. Therefore, heart disease diagnosis is found to have achieved huge focus all over the world among the healthcare sector. Data mining algorithms played a significant role in heart disease diagnosis with good efficiency. To overcome the above-mentioned problems, EGA (Enhanced Genetic Algorithm) based FWSVM (Fuzzy Weight updating Support Vector Machine) algorithm is proposed. Three major phases are such as preprocessing, feature section and classification. Initially, data is pre-processed for enhanced classification accuracy using KMC (K-Means Clustering) followed by feature selections where EGA algorithm is used. Finally, classification if heart diseases is done using FWSVM algorithm introduces fuzzy weight values to compute and validate the input data (e.g., patient id, age, sex, angina, smoking and so on) to predict earlier stage of disease. The experimental outcome concluded that the proposed EGA + FWSVM algorithm provides better heart disease detection performance for the given heart disease dataset. The performance evaluation metrics are accuracy, recall, precision, and f-measure, specificity and time complexity values proposed compare than existing techniques.

Keywords

• Heart disease data
• K-means clustering (KMC) algorithm
• Enhanced genetic algorithm (EGA)
• Fuzzy weight updating SVM (FWSVM) algorithm
• Prediction results