Table 12 Results of classifiers (in %) with MAD.

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

69

71

75

71

71

73

75

DT

76.7

71.4

73

76

66

74

76

RF

79.3

76.5

78

80

72

79

83

KNN

78.7

71

75

72

70

73

77

SVM

68.9

69.3

74

69

70

71

75

GNB

70

68.9

76

63

76

69

76

XGBoost

99.8

81.9

85

82

82

83

88

AdaBoost

75.3

72.7

78

70

76

74

78

SGD

56.7

58

57

98

10

72

54

GB

97.2

77.7

80

80

76

80

86

ETC

74.1

71.8

75

73

71

94

79

CatBoost

76.7

74

92

58

94

71

89

LightGBM

90.8

73.1

89

58

92

70

88

MLP

70.9

68.9

78

60

79

68

78

RNN

75

71

88

34

94

49

64

LSTM

74.1

72.7

86

37

93

52

65

GRU

74.5

72.3

87

31

94

46

63

Bi-LSTM

72.4

71.4

88

35

94

50

65

Bi-GRU

74.6

73.1

90

29

96

44

63

CNN

74.1

71.9

87

34

94

49

67

Hybrid Model

74.1

70.6

85

36

93

51

79

  1. Significant values are in [bold].