Skip to main content
Fig. 4 | BioData Mining

Fig. 4

From: Gene function finding through cross-organism ensemble learning

Fig. 4

Prediction model precision vs. number of predictions. Precision (X axis) and number of predictions (Y axis) of prediction models trained on Homosapiens known annotations, either excluding or including IEA-only annotations, and tested on four target organisms (one per column of plots) on each of the considered ensemble methods (one per row of plots). Each point in the plots represents a test result obtained with a different value of the predicted annotation likelihood threshold ρ, ranging from 0.05 to 0.95; leftmost points in each plot correspond to values lower than ρ

Back to article page
\