Abstract:

There is certain scientific interest in the field of process mining-based quality assessment and prediction models (using selected AI methods) that support faultless manufacturing. In the case of solving complex issues regarding such multidimensional problems it is advisable to search for new models that integrate several intelligent methods, e.g. linear regression for performing feature selection and Fully Connected Deep Neural Networks (FCDNN) for classifying business process instances. The purpose of this paper is to propose a quality assessment model based on mixed intelligent methods (MIM) that can be used for manufacturing process mining. Theoretical considerations are illustrated with an example model implementation and results of computational experiments on industrial data. Conducting study on the use of MIM in process mining may be a development impulse for further research on quality assessment and prediction in Business Process Management (BPM) field as well as business and management study.