However, existing feature extraction methods might include laborious processes and it is hard for them to minimize the irrelevant information while preserving the important information. To solve this ...

In this research, we propose a generic deep autoencoder model for automated feature extraction from the raw sensor data. Extracted deep features are classified with random forest algorithm for fault ...