Table 5 Classifier performance metrics (80% training, 20% testing).

From: Enhanced sensing performance through the integration of denoising autoencoder and ensembling techniques

Classifier

Accuracy (%)

F1-Score (%)

MCC (%)

Training

Testing

Training

Testing

Training

Testing

DT

96.99

96.93

97.03

96.72

95.98

95.35

GNB

95.94

95.95

95.47

95.64

95.92

95.25

NN

96.73

96.76

96.39

96.44

95.75

95.85

Proposed EC

99.66

99.13

99.60

99.03

99.16

98.00

RUSBoost

99.37

99.03

99.24

98.94

98.44

97.80

Bagged

99.53

99.47

99.43

99.27

98.82

98.50

RFC

96.63

96.76

96.67

96.44

96.32

95.85

KNN

96.57

96.60

96.39

96.31

95.99

95.58

  1. All valus rounded to 2 decimal places; DT: Decision Tree; GNB: Gaussian Naive Bayes; NN: Neural Network; EC: Ensemble Classifier; RFC: Random Forest Classifier; KNN: K-Nearest Neighbors.
Back to article page