Explainable multi-class anomaly detection on functional data
This work addresses anomaly detection for functional data in industrial applications, but it is incremental as it combines existing methods.
The paper tackles the problem of detecting and explaining anomalies in multivariate functional data by transforming series into features and using an Isolation forest algorithm, with explainability achieved through SHAP coefficients and a supervised decision tree, applied on simulated and real industrial data.
In this paper we describe an approach for anomaly detection and its explainability in multivariate functional data. The anomaly detection procedure consists of transforming the series into a vector of features and using an Isolation forest algorithm. The explainable procedure is based on the computation of the SHAP coefficients and on the use of a supervised decision tree. We apply it on simulated data to measure the performance of our method and on real data coming from industry.