These graphs show the aggregated classification over three different runs, where for every category (FN, TP, TN, FP) the minimal number over the different runs is taken. For example, if 50 traces are TP in the first run, 45 in the second, and 48 in the third, then the value 45 is used as aggregated value. As a result, these graphs can show discrepancies between the classifications in different runs.
For the Inductive Miner, there is a discrepancy for pdc_2019_10.xes. Although for this log some of the traces fail to replay within 10 seconds, which could also explain the discrepancy, the nets as discovered by the second run does differ from the net as discovered by the other two runs.