Deep Autoencoders with Value-at-Risk Thresholding for Unsupervised Anomaly Detection
This work addresses anomaly detection for real-world monitoring and surveillance applications, but appears incremental as it builds on existing deep autoencoder approaches with a new thresholding method.
The paper tackles unsupervised anomaly detection in streaming data by training a deep autoencoder model and using a novel value-at-risk thresholding mechanism for detection. It reports results on changedetection net, showing comparisons against subspace methods, though no concrete numbers are provided in the abstract.
Many real-world monitoring and surveillance applications require non-trivial anomaly detection to be run in the streaming model. We consider an incremental-learning approach, wherein a deep-autoencoding (DAE) model of what is normal is trained and used to detect anomalies at the same time. In the detection of anomalies, we utilise a novel thresholding mechanism, based on value at risk (VaR). We compare the resulting convolutional neural network (CNN) against a number of subspace methods, and present results on changedetection net.