Table 18 Results of classifiers (in %) with PCA.

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

77.4

78.6

79

83

73

81

83

DT

83.2

78.6

72

49

79

80

83

RF

80.7

80.7

81

85

76

83

88

KNN

82.3

74.8

76

79

69

78

80

SVM

79

78.1

78

84

71

81

83

GNB

77.4

77.3

77

83

70

80

83

XGBoost

99.6

83.6

87

82

85

85

91

AdaBoost

81

76.9

79

79

74

79

86

SGD

70.7

68.9

84

54

87

66

81

GB

96.9

84.5

86

86

82

86

90

ETC

80.9

79.8

82

81

79

82

87

CatBoost

81.6

75.6

95

59

96

73

92

LightGBM

82.5

76.1

93

61

94

74

92

MLP

76.5

75.6

82

72

80

76

82

RNN

79.6

77.3

86

45

91

59

68

LSTM

82.9

79.8

87

51

91

64

71

GRU

82.4

79.4

85

43

90

58

67

Bi-LSTM

82.8

80.3

85

46

90

60

68

Bi-GRU

79

76.5

87

80

91

63

70

CNN

81.3

78.8

86

52

91

61

72

Hybrid Model

81.3

79.4

87

45

92

59

87

  1. Significant values are in [bold].