Mining aberrant driving behavior characteristicsMining aberrant driving behavior characteristics Σύντομη Δημοσίευση σε Συνέδριο Conference Short Paper 2015-11-052005enThis article demonstrates the contribution of machine learning in modeling of aberrant driving characteristics. We demonstrate how machine learning can be deployed to identify, classify, and model interesting and useful knowledge patterns, which are hidden in self reported replies to a Driving Behavior Questionnaire. Demonstration is based on empirical results drawn from 1453 responders in Greece. Experimentation was based on two machine learning approaches: decision tree learning and association rule mining and learning focused on the exact driver characteristics thought to be responsible for an accident to occur. Accuracy of results was assessed using randomized testinghttp://creativecommons.org/licenses/by/4.0/Διεθνές συνέδριο WSEAS  Moustakis Vasilis Μουστακης Βασιλης Tsironis Loukas Τσιρωνης Λουκας Kontogiannis Thomas Κοντογιαννης Θωμας H. Mavropoulos