A Predictive Validation Method for Discovering Discrete Event Simulation Models

System inference represents a challenging issue in system theory. To deal with system inference, D2FD (Data to Fuzzy-DEVS) method has been proposed to discover a Fuzzy-DEVS model from event data. This method has been successfully implemented on the process mining framework (ProM). The validation of this method is able to use a morphism-based model approximation at the replicative level. However, there is a need of predictive validation method for emergency. In this paper, we propose a predictive validation method for D2FD method integrated with granger causality. This method is able to generate a new business process model which is causally influenced by its underlying factors. A predictive model is shown in the real case study from Italian University.