Random forest of oblique decision trees for ERP semi-automatic configuration

Abstract

Enterprise Resource Planning (ERP) is one of the most important parts of company’s information system. However few ERP implementation projects are delivered on time. Configuration of ERP based on questionnaires and/or interviews is time consuming and expensive, especially because many answers should be checked and corrected by ERP consultants. Supervised learning algorithms can thus be useful to automatically detect wrong and correct answers. Comparison done on real free open-source ERP data shows that random forest of oblique decision trees is very efficient.