Fig 1.
System architecture of this study.
Table 1.
The dataset’s summarized description.
Fig 2.
Histogram of attributes.
Fig 3.
Feature selection procedure.
Fig 4.
Illustration of all features correlation.
Table 2.
The summarized output of the eight model for the important feature selection.
Table 3.
Description of the train test split dataset.
Table 4.
Hyperparameters of machine learning algorithms used grid search.
Table 5.
Overall performance of all of the models summed up cross validation with grid search.
Fig 5.
Graphical representation of all models confusion matrix.
Fig 6.
The ROC curve for the experiment.
Table 6.
Overall performance of all of the models summed up without cross validation and grid search.
Fig 7.
(a) XGBoost Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (b) RF Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (c) KNN Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (d) SVM Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (e) ABR Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (f) NB-Gaussian Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (g) LR Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (h) Linear Regression Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search; (i) DT Model performance on Cross Validation with Grid, Cross Validation without Grid Search, Without Cross validation and Grid Search.
Table 7.
Overall performance of all of the models summed up cross validation without grid search.
Table 8.
The summarized comparison of the heart disease prediction model’s performance.
Fig 8.
The illustration of proposed clinical application.