Explainable multi-class anomaly detection on functional data

Mathieu Cura; Katarina Firdova; Céline Labart; Arthur Martel

Explainable multi-class anomaly detection on functional data

Machine Learning 2022-05-09 v1 Machine Learning Statistics Theory Statistics Theory

Authors: Mathieu Cura , Katarina Firdova , Céline Labart , Arthur Martel

Abstract

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.

Keywords

tree-based methods functional data analysis machine learning

Cite

@article{arxiv.2205.02935,
  title  = {Explainable multi-class anomaly detection on functional data},
  author = {Mathieu Cura and Katarina Firdova and Céline Labart and Arthur Martel},
  journal= {arXiv preprint arXiv:2205.02935},
  year   = {2022}
}

Related papers

View all related →