Estimation of Dimensions Contributing to Detected Anomalies with Variational Autoencoders
This work addresses the need for interpretable anomaly detection in multidimensional data monitoring, offering a domain-specific improvement over existing methods.
The paper tackles the problem of poor interpretability in deep learning-based anomaly detection by proposing a novel algorithm that uses variational autoencoders to estimate which dimensions contribute to detected anomalies, with experiments on benchmark datasets showing it extracts these dimensions more accurately than baseline methods.
Anomaly detection using dimensionality reduction has been an essential technique for monitoring multidimensional data. Although deep learning-based methods have been well studied for their remarkable detection performance, their interpretability is still a problem. In this paper, we propose a novel algorithm for estimating the dimensions contributing to the detected anomalies by using variational autoencoders (VAEs). Our algorithm is based on an approximative probabilistic model that considers the existence of anomalies in the data, and by maximizing the log-likelihood, we estimate which dimensions contribute to determining data as an anomaly. The experiments results with benchmark datasets show that our algorithm extracts the contributing dimensions more accurately than baseline methods.