2023 A Study on the Performance Improvement of Rotational Machinery Anomaly Detection through Deep Feature Ensemble of Self-Attention LSTM Autoencoder and LSTM Autoencoder Based on Unsupervised Learning