A machine learning approach to the automatic synthesis of mechanistic knowledge for engineering decision-making

Inductive learning is proposed as a tool for synthesizing domain knowledge from data generated from a model-based simulator. To use an inductive engine to generate decision rules, a preclassification process is necessary in the presence of multiple competing objectives. Instead of relying on a domain expert to perform this preclassification, a clustering algorithm is used to eliminate the human bias involved in the selection of a classification function for the preclassification. It is shown that the use of a clustering algorithm for preclassification not only further automates the process of knowledge synthesizing, but also improves the quality of the rules generated by the inductive engine