In this study the applicability of autoencoder-based deep neural networks to solve fault classification problems using vibration signals. This work will aim to construct CNN and LSTM type ...

The D-CNN-LSTM Autoencoder method optimizes the anomaly detection rate for all of the anomalies, specifically in the case of low magnitude anomalies, enhancing F1-score up to 18.12% in single types of ...