A Goal Programming-Based Methodology for Machine Learning Model Selection Decisions: A Predictive Maintenance Application

The paper develops a goal programming-based multi-criteria methodology, for assessing different machine learning (ML) regression models under accuracy and time efficiency criteria. developed methodology provides users with high flexibility in the as it allows fast computationally efficient sensitivity analysis of significance weights well threshold values. Four were assessed, namely decision tree, random forest, support vector neural network. was employed to forecast failures NASA Turbofans. results reveal that tree (DTR) seems be preferred low values (up 30%) As tend increase higher values, forest (RFR) best choice. preference RFR model however, change towards adoption network equal than 90%.