Fig. 3 | Scientific Reports

Fig. 3

From: Evaluating the factors influencing accuracy, interpretability, and reproducibility in the use of machine learning classifiers in biology to enable standardization

Fig. 3

Impacts of data type on model accuracy and feature importance between classifiers. (A) Testing accuracy of five classifiers over 500 random training and testing data splits for the transcript (blue) and protein (yellow) data. Solid line: mean accuracy on the hold out test set for the 500 training interactions, Ribbons: one standard deviation above and below the mean. (B) Scaled importance values closest to 1 indicate highest importance for a given classifier. Widest portions of the distribution indicate the most frequent scaled importance across the fifty repetitions of five-fold cross validation feature selection for each of the models.

Back to article page