Multi-Objective Optimization of Performance and Interpretability of Tabular Supervised Machine Learning Models