Unsupervised Anomaly Detection on Temporal Multiway Data
This addresses anomaly detection for temporal multiway data, which is less studied, but it appears incremental as it builds on existing matrix-native recurrent neural networks.
The paper tackled unsupervised anomaly detection on temporal multiway data, such as matrices observed over time, by exploring strategies like compressing-decompressing and encoding-predicting using matrix LSTMs, finding that encoding-predicting works well due to compactness and better data fit, with specific insights like matrix LSTMs compressing noisy data near perfectly.
Temporal anomaly detection looks for irregularities over space-time. Unsupervised temporal models employed thus far typically work on sequences of feature vectors, and much less on temporal multiway data. We focus our investigation on two-way data, in which a data matrix is observed at each time step. Leveraging recent advances in matrix-native recurrent neural networks, we investigated strategies for data arrangement and unsupervised training for temporal multiway anomaly detection. These include compressing-decompressing, encoding-predicting, and temporal data differencing. We conducted a comprehensive suite of experiments to evaluate model behaviors under various settings on synthetic data, moving digits, and ECG recordings. We found interesting phenomena not previously reported. These include the capacity of the compact matrix LSTM to compress noisy data near perfectly, making the strategy of compressing-decompressing data ill-suited for anomaly detection under the noise. Also, long sequence of vectors can be addressed directly by matrix models that allow very long context and multiple step prediction. Overall, the encoding-predicting strategy works very well for the matrix LSTMs in the conducted experiments, thanks to its compactness and better fit to the data dynamics.