Table 19 Performance of proposed model and state-of-the-art models.

From: An extensive experimental analysis for heart disease prediction using artificial intelligence techniques

Refs.

Dataset

Model

Performance metric (in %)

Accuracy

Precision

Sensitivity

Specificity

F1 Score

AUC

5

Cleveland

DNN

93.33

87.8

91.83

-

94

6

Cleveland,

Hungarian,

Switzerland,

Long Beach VA

DNN

83.03

90.9

69.27

87.37

12

Cleveland

HRFLM

88.4

90.1

92.8

82.6

90

-

14

Cleveland

RF + FAMD

93.44

89.28

96.96

92.59

93.12

15

Heart-failure-

clinical-records-

dataset

KNN

90.78

26

Heart-failure-

clinical-records-

dataset

ETC

92.62

93

93

93

27

Cleveland

FCMIM + SVM

92.37

89

98

-

28

Z-Alizadeh

Sani dataset

GNB

95.43

95.84

94.44

96.77

Statlog

GNB

93.3

89.2

96.7

92.1

Cardiovascular

disease dataset

GNB

73.2

69.3

77

71.9

29

CHD dataset

LightGBM

93

96.3

89.7

96.3

92.9

97.8

Proposed

Heart disease

Dataset

(Comprehensive)

XGBoost

97.3

97

98

98

98

98

  

Superior model

XGBoost

XGBoost

XGBoost

XGBoost,

FCMIM + SVM

XGBoost

XGBoost

  1. Significant values are in [bold].