Figure 6.
The number of training data sets used plotted against the test set performance of every algorithm: (a) Accuracy, (b) Area under the ROC-curve (AUC). The error bars show the 95% confidence interval.
The number of training data sets used plotted against the test set performance of every algorithm: (a) Accuracy, (b) Area under the ROC-curve (AUC). The error bars show the 95% confidence interval.