Fig. 7

Cross-dataset classification results. Each cell (i,j) represents the mean accuracy of the 20 XGBoost classifiers that were trained on dataset i (in "Evaluation of different machine-learning methods" section) and tested on dataset j (ACC(i, j)). The black frames indicate the results of dataset pairs originating from the same species. The accuracy results for pairs (i,i) were taken from "Intra-dataset analysis" section. Note that, for the ease of the interpretation of the results, the color scale is inverse to the scale used for the KL-divergence plot in Fig. 5