Fig. 4
From: kMoL: an open-source machine and federated learning library for drug discovery
Results of federated learning classification experiments: The X-axis represents the identifier of the experiment, and the Y-axis represents the \(ROC\text {-}AUC\) metric value. The client experiment settings vary in terms of the number of clients among which the data is distributed. The epoch experiment settings vary in terms of the number of epochs trained per round. The imbalance experiment settings vary in terms of the ratios of unevenly distributed data. (e.g., I-60–40 represents the setting where one client receives 60%, and the other client receives 40% of the data samples)