Challenging the Performance-Interpretability Trade-off: An Evaluation of Interpretable Machine Learning Models