Improving Interpretability of Scores in Anomaly Detection Based on Gaussian-Bernoulli Restricted Boltzmann Machine
This work addresses interpretability issues in anomaly detection for practitioners using GBRBMs, though it is incremental as it builds on existing methods.
The study tackled the problem of setting classification thresholds in Gaussian-Bernoulli restricted Boltzmann machine-based anomaly detection by proposing a measure to improve score interpretability based on its cumulative distribution, establishing a guideline for threshold setting validated through numerical experiments.
Gaussian-Bernoulli restricted Boltzmann machines (GBRBMs) are often used for semi-supervised anomaly detection, where they are trained using only normal data points. In GBRBM-based anomaly detection, normal and anomalous data are classified based on a score that is identical to an energy function of the marginal GBRBM. However, the classification threshold is difficult to set to an appropriate value, as this score cannot be interpreted. In this study, we propose a measure that improves score's interpretability based on its cumulative distribution, and establish a guideline for setting the threshold using the interpretable measure. The results of numerical experiments show that the guideline is reasonable when setting the threshold solely using normal data points. Moreover, because identifying the measure involves computationally infeasible evaluation of the minimum score value, we also propose an evaluation method for the minimum score based on simulated annealing, which is widely used for optimization problems. The proposed evaluation method was also validated using numerical experiments.