Unsupervised Transfer Learning for Anomaly Detection: Application to Complementary Operating Condition Transfer

TL;DR

This paper introduces an unsupervised transfer learning framework that aligns data distributions from different units to improve anomaly detection in changing industrial operating conditions, using adversarial deep learning and a novel loss function.

Contribution

It presents a novel unsupervised transfer learning method for one-class classification that transfers complementary data between units to enhance anomaly detection.

Findings

Improved anomaly detection accuracy on three open source datasets.

Effective alignment of data distributions from different units.

Enhanced robustness of detectors in varying operating conditions.

Abstract

Anomaly Detectors are trained on healthy operating condition data and raise an alarm when the measured samples deviate from the training data distribution. This means that the samples used to train the model should be sufficient in quantity and representative of the healthy operating conditions. But for industrial systems subject to changing operating conditions, acquiring such comprehensive sets of samples requires a long collection period and delay the point at which the anomaly detector can be trained and put in operation. A solution to this problem is to perform unsupervised transfer learning (UTL), to transfer complementary data between different units. In the literature however, UTL aims at finding common structure between the datasets, to perform clustering or dimensionality reduction. Yet, the task of transferring and combining complementary training data has not been studied. Our proposed framework is designed to transfer complementary operating conditions between different units in a completely unsupervised way to train more robust anomaly detectors. It differs, thereby, from other unsupervised transfer learning works as it focuses on a one-class classification problem. The proposed methodology enables to detect anomalies in operating conditions only experienced by other units. The proposed end-to-end framework uses adversarial deep learning to ensure alignment of the different units' distributions. The framework introduces a new loss, inspired by a dimensionality reduction tool, to enforce the conservation of the inherent variability of each dataset, and uses state-of-the art once-class approach to detect anomalies. We demonstrate the benefit of the proposed framework using three open source datasets.