 This paper proposes a novel methodology for classifying two types of rotary system faults, unbalance and shaft bow, based on the extracted features from a wavelet time-scattering, WTS, algorithm. The proposed methodology was evaluated using a long short-term memory, LSTM, neural network and a support vector machine SVM. The results showed that the WTS-based approach outperformed both the LSTM and SVM models in terms of accuracy and precision. This article was authored by Nima Rezizada, Mario de Oliveira, Donato Perfetto, and others.