Table 2.

Evaluation of the 15 classifiers. Fifteen classifiers were obtained by training three algorithms with five different training sets. The performance of each classifier was evaluated using a test set conformed by genes that were not used during training. The table shows the mean and standard deviation of the accuracy, the F1 score and the area under the ROC curve of the five classifiers trained with each algorithm. The last three rows show the area under the ROC obtained by other colleagues when predicting other biological functions through machine learning