Fig. 1

A map presenting the origin of the samples in the MetaSUB dataset. The eight cities marked with blue color are included in the primary dataset, and four cities marked with red color are the origins of the samples included in the C2 and C3 sets. On the map, we show the classification accuracies (obtained using the proposed method) for the cities from the primary dataset—blue indicates the scores for the primary dataset (based on leave-one-out cross validation), and green shows the scores for the C1 set (which includes samples from four cities out of eight from the primary dataset)