Table 17 Results of classifiers (in %) with LDA.

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

Classifier

Training accuracy

Testing accuracy

Precision

Sensitivity

Specificity

F1 score

AUC

LR

83.7

85.7

88

86

85

87

90

DT

84.9

87

87

90

83

88

92

RF

85.2

87

87

90

83

88

93

KNN

83.4

84.9

85

89

80

87

84

SVM

83.1

86.6

87

89

84

88

90

GNB

81.3

83.2

86

82

84

84

89

XGBoost

85.1

87.8

88

91

84

89

92

AdaBoost

83.7

87

86

91

82

89

92

SGD

83.5

84.5

86

86

83

86

89

GB

85.5

87

87

90

83

88

92

ETC

84.7

84.9

88

84

86

86

92

CatBoost

84.9

79

91

69

92

78

92

LightGBM

84.8

79

91

69

92

78

92

MLP

85.7

85.7

86

89

82

87

93

RNN

85

87

90

73

90

80

81

LSTM

85.3

87

93

66

94

77

80

GRU

84.8

87.4

92

64

94

76

79

Bi-LSTM

85.4

85.7

92

64

94

76

79

Bi-GRU

85.9

87.4

94

61

94

74

78

CNN

85.4

87.1

92

67

93

77

82

Hybrid Model

85.3

87.8

89

71

90

79

92

  1. Significant values are in [bold].