TUM Logo

R2-AD2: Detecting Anomalies by Analysing the Raw Gradient

Neural networks follow a gradient-based learning scheme, adapting their mapping parameters by back-propagating the output loss. Samples unlike the ones seen during training cause a different gradient distribution. Based on this intuition, we design a novel semi-supervised anomaly detection method called R2-AD2. By analysing the temporal distribution of the gradient over multiple training steps, we reliably detect point anomalies in strict semi-supervised settings. Instead of domain dependent features, we input the raw gradient caused by the sample under test to an end-to-end recurrent neural network architecture. R2-AD2 works in a purely data-driven way, thus is readily applicable in a variety of important use cases of anomaly detection.

R2-AD2: Detecting Anomalies by Analysing the Raw Gradient

Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2022, Grenoble, France, September 19–23, 2022, Proceedings, Part I

Authors: Jan-Philipp Schulze, Philip Sperl, Ana Radutoiu, Carla Sagebiel, and Konstantin Böttinger
Year/month: 2023/3
Pages: 209 - 224
Fulltext: click here

Abstract

Neural networks follow a gradient-based learning scheme, adapting their mapping parameters by back-propagating the output loss. Samples unlike the ones seen during training cause a different gradient distribution. Based on this intuition, we design a novel semi-supervised anomaly detection method called R2-AD2. By analysing the temporal distribution of the gradient over multiple training steps, we reliably detect point anomalies in strict semi-supervised settings. Instead of domain dependent features, we input the raw gradient caused by the sample under test to an end-to-end recurrent neural network architecture. R2-AD2 works in a purely data-driven way, thus is readily applicable in a variety of important use cases of anomaly detection.

Bibtex:

@article {
author = { Jan-Philipp Schulze and Philip Sperl and Ana Radutoiu and Carla Sagebiel and Konstantin Böttinger},
title = { R2-AD2: Detecting Anomalies by Analysing the Raw Gradient },
journal = { Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2022, Grenoble, France, September 19–23, 2022, Proceedings, Part I },
year = { 2023 },
month = { March },
pages = { 209 - 224 },
url = { https://doi.org/10.1007/978-3-031-26387-3_13 },

}