Process Predictive Analytics: A Comparison of Approaches

Predictive process analytics consist in combining the power of process mining and data science to foresee the future outcome of instances of business processes, where the outcome is measured via customizable Key Performance Indicators (KPIs). This thesis focuses on object-centric processes, which implement an innovative paradigm where an instance of one process is not executed in isolation but interacts with other instances of the same or other processes. Interactions take place through bridging events where instances of possibly different processes synchronize. This thesis explores and compares various types of prediction models applicable in predictive analytics for object-centric processes, providing a comprehensive understanding of their strengths, limitations, and practical applications. The thesis’ work started by discussing the theoretical foundations behind the approach, and continued by applying the theoretical principles to our proposed framework, completing by comparing the performances of alternative techniques.