DTOR: Decision Tree Outlier Regressor to explain anomalies
This addresses the need for actionable explanations of anomalies in domains like fraud detection, though it is incremental as it builds on existing rule-based approaches.
The paper tackles the problem of explaining anomalies detected by machine learning models, presenting DTOR, a method that uses decision tree regressors to generate rule-based explanations for individual data points, with results showing robustness in high-dimensional datasets and comparable performance to Anchors while reducing execution time.
Explaining outliers occurrence and mechanism of their occurrence can be extremely important in a variety of domains. Malfunctions, frauds, threats, in addition to being correctly identified, oftentimes need a valid explanation in order to effectively perform actionable counteracts. The ever more widespread use of sophisticated Machine Learning approach to identify anomalies make such explanations more challenging. We present the Decision Tree Outlier Regressor (DTOR), a technique for producing rule-based explanations for individual data points by estimating anomaly scores generated by an anomaly detection model. This is accomplished by first applying a Decision Tree Regressor, which computes the estimation score, and then extracting the relative path associated with the data point score. Our results demonstrate the robustness of DTOR even in datasets with a large number of features. Additionally, in contrast to other rule-based approaches, the generated rules are consistently satisfied by the points to be explained. Furthermore, our evaluation metrics indicate comparable performance to Anchors in outlier explanation tasks, with reduced execution time.