Skip to main content

Table 5 Performance results based on test data using four classifiers: KNN, logistic regression, SVM, random forest

From: Automated chest screening based on a hybrid model of transfer learning and convolutional sparse denoising autoencoder

Classifier Data augmentation Precision Recall F1 AUC
KNN Without augmentation 0.73 0.58 0.64 0.753
0.87 0.93 0.9
SMOTE 0.44 0.84 0.58 0.752
0.93 0.67 0.78
Positive augmentation 0.57 0.68 0.62 0.758
0.89 0.68 0.87
4× augmentation 0.57 0.84 0.62 0.758
0.89 0.89 0.87
Logistic Without augmentation 0.76 0.54 0.63 0.744
0.87 0.95 0.91
SMOTE 0.61 0.82 0.64 0.79
0.92 0.76 0.84
Positive augmentation 0.57 0.66 0.61 0.754
0.89 0.84 0.86
4× augmentation 0.63 0.72 0.67 0.792
0.91 0.87 0.89
SVM Without augmentation 0.63 0.73 0.68 0.798
0.91 0.86 0.89
SMOTE 0.63 0.77 0.69 0.813
0.92 0.86 0.89
Positive augmentation 0.59 0.64 0.61 0.75
0.88 0.86 0.87
4× augmentation 0.61 0.71 0.66 0.784
0.9 0.86 0.88
Random forest Without augmentation 0.69 0.38 0.49 0.663
0.83 0.95 0.88
SMOTE 0.62 0.5 0.56 0.704
0.85 0.9 0.88
Positive augmentation 0.54 0.63 0.58 0.73
0.88 0.83 0.85
4× augmentation 0.59 0.54 0.56 0.71
0.86 0.88 0.87