A Hybrid Intelligent System Framework for the Prediction of Heart Disease Using Machine Learning Algorithms

Table 6

10-fold CV classification performance evaluation of different classifiers on Cleveland heart disease dataset on full features.

Predictive model	Classifiers performance evaluation metrics
	Accuracy (%)	Specificity (%)	Sensitivity (%)	MCC	AUC (%)	Processing time (s)
Logistic regression (C = 10)	84	85	83	89	84	19.213
K-nearest neighbor (K-NN, K = 9)	76	74	73	76	73	29.400
Artificial neural network (13, 16, 2)	74	73	74	50	69	21.600
SVM (kernel = RBF, C = 100, = 0.0001)	86	88	78	85	86	15.234
SVM (kernel = linear)	75	78	75	78	74	18.239
Naive Bayes	83	87	78	80	84	34.101
Decision tree	74	76	68	75	76	21.911
Random forest (100)	83	70	94	82	83	15.121