Table 1 Performance metrics of Linear SVM classi er assessed by standard 10-fold cross validation, leave-one-outbreak-out validation and random undersampling methods
From: Analysis of heterogeneous genomic samples using image normalization and machine learning
Sampling Methods | Accuracy | Precision-Chronic | Precision-Recent | Recall-Chronic | Recall-Recent | AUC |
|---|---|---|---|---|---|---|
Standard 10-fold cross-validation | 97.545% | 98.105% | 96.515% | 98.446% | 95.364% | 96.905% |
Leave-one-outbreak-out | 96.075% | 97.004% | 91.0% | 98.446% | 83.5% | 90.973% |
Random undersampling | 95.164% | 96.328% | 94.661% | 94.155% | 96.173% | 95.164% |