 This paper proposes a new approach for bearing fault diagnosis using multi-scale permutation entropy, MPE, which is more effective than existing methods such as single-scale permutation entropy, PE, and multi-scale entropy, MSE. The MPE algorithm extracts features from vibrations caused by bearing faults, and then uses support vector machines, SVM, to classify these features into different categories corresponding to various types of bearing faults. This method outperformed other approaches in terms of accuracy and robustness. This article was authored by Chen Jian-Ding, Chen Qi-Wang, Qi Yuan-Wu, and others.