We compare the models from the commonly used CNN or LSTM-based autoencoder anomaly diagnosis technique to its variant model DAGMM (Deep Autoencoding Gaussian Mixture model), deep learning-based SVDD ...

On this basis, we propose the denoising autoencoder based on the broad learning system (DBLS-AE), which sufficiently learns the anomalous patterns, achieving efficient anomaly detection with low ...