You are here
Everyday life routinely poses questions and challenges on prediction: “how much longer is my 10 years old cat expected to live?”; “will the 12:00 noon fast train from Rome to Milan arrive on time so that I can catch my connection?”; “what will the share price of my investment be in 10 days?”. For all these questions, asking whether in the future a certain event will occur or not (a certain property will be satisfied or not), we cannot expect to provide a true answer based on what we already know. Yet, what we learned from history and past experiences often suggests a reasonable answer.
Analogous questions and predictive challenges can arise during the execution of business processes. For example, in a medical process execution a doctor may ponder whether a surgery, a pharmacological therapy or a manipulation is the best choice to be made in order to guarantee the patient recovery. Predictive business process monitoring is a family of techniques that apply what we do in everyday life to the field of business processes. In particular, predictive process monitoring exploits event logs, which are more and more widespread in modern information systems, to predict how current (uncompleted) cases will unfold up to their completion.
ThePreMo tool is a predictive process monitoring framework for estimating the likelyhood that an ongoing case will lead to a certain outcome among a set of possible outcomes. It includes a range of techniques to adapto to different scenarios as well as hyperparameter optimization methods to select a suitable framework instance for a given dataset.