Process Predictive Monitoring Based on Federated Learning: Challenges and Architecture Proposal
Wanting Yin, Yixin Sun, Xinda Li, Rongjia Song · 2023
Process predictive monitoring refers to prediction the future activities, execution time, resources and other attribute in the ongoing business process. Aiming at the problem that the data scale of a single enterprise or organization is limited in business process, the learning model is difficult to further improve the prediction ability. This paper proposes a novel process predictive monitoring architecture proposal based on federated learning, and applies federated learning in process predictive monitoring to avoid the original data transmission and protect the privacy of data. Through aggregation training, the diversity of data is increased and the generalization ability and prediction performance of the model are improved.