An Optimization Framework Using Flood Algorithm-Based Feature Selection for Heart Disease Prediction

Abstract

Heart disease is one of the common disease  and has  been ranked as the one most prevalent. Early detection is the important step to reduce mortality rates and treatment costs.  This research proposes a technique for extracting the most relevant features using flood algorithm (FLA)  to improve the performance of the diagnosis model. The feature selection model is evaluated using parameters derived from the confusion matrix and using six models:MLP, TabNet,  Random Forest, SVM, Logistic Regression, and XGBoost, the models were testing to the heart disease dataset. The best results for this models were obtained using XGBoost  with 7 selected features from 13features and the resulting performance for accuracy  was 98.9%, specificity 98.8%, sensitivity 99.00%, area under the curve (AUC) 99.8%, and F1-Score 99%.