Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantum Autoencoder for Multivariate Time Series Anomaly Detection

Kilian Tscharke, Maximilian Wendlinger, Afrae Ahouzi, Pallavi Bhardwaj, Kaweh Amoi-Taleghani, Michael Schrödl-Baumann, Pascal Debus

TL;DR

Conventional AD for high-dimensional MTS in enterprise telemetry is challenged by scale and labeled anomaly scarcity. The authors develop a Quantum Autoencoder (QAE) architecture tailored for MTS AD, using data re-upload encoding and a trash-qubit measurement to derive anomaly scores without requiring labeled anomalies. Experiments show the QAE is competitive with neural-network autoencoders while using substantially fewer trainable parameters, across SMD, Pasta, and MSCM datasets. This work demonstrates a viable quantum-enhanced, semisupervised AD approach for real-world enterprise observability pipelines.

Abstract

Anomaly Detection (AD) defines the task of identifying observations or events that deviate from typical - or normal - patterns, a critical capability in IT security for recognizing incidents such as system misconfigurations, malware infections, or cyberattacks. In enterprise environments like SAP HANA Cloud systems, this task often involves monitoring high-dimensional, multivariate time series (MTS) derived from telemetry and log data. With the advent of quantum machine learning offering efficient calculations in high-dimensional latent spaces, many avenues open for dealing with such complex data. One approach is the Quantum Autoencoder (QAE), an emerging and promising method with potential for application in both data compression and AD. However, prior applications of QAEs to time series AD have been restricted to univariate data, limiting their relevance for real-world enterprise systems. In this work, we introduce a novel QAE-based framework designed specifically for MTS AD towards enterprise scale. We theoretically develop and experimentally validate the architecture, demonstrating that our QAE achieves performance competitive with neural-network-based autoencoders while requiring fewer trainable parameters. We evaluate our model on datasets that closely reflect SAP system telemetry and show that the proposed QAE is a viable and efficient alternative for semisupervised AD in real-world enterprise settings.

Quantum Autoencoder for Multivariate Time Series Anomaly Detection

TL;DR

Conventional AD for high-dimensional MTS in enterprise telemetry is challenged by scale and labeled anomaly scarcity. The authors develop a Quantum Autoencoder (QAE) architecture tailored for MTS AD, using data re-upload encoding and a trash-qubit measurement to derive anomaly scores without requiring labeled anomalies. Experiments show the QAE is competitive with neural-network autoencoders while using substantially fewer trainable parameters, across SMD, Pasta, and MSCM datasets. This work demonstrates a viable quantum-enhanced, semisupervised AD approach for real-world enterprise observability pipelines.

Abstract

Anomaly Detection (AD) defines the task of identifying observations or events that deviate from typical - or normal - patterns, a critical capability in IT security for recognizing incidents such as system misconfigurations, malware infections, or cyberattacks. In enterprise environments like SAP HANA Cloud systems, this task often involves monitoring high-dimensional, multivariate time series (MTS) derived from telemetry and log data. With the advent of quantum machine learning offering efficient calculations in high-dimensional latent spaces, many avenues open for dealing with such complex data. One approach is the Quantum Autoencoder (QAE), an emerging and promising method with potential for application in both data compression and AD. However, prior applications of QAEs to time series AD have been restricted to univariate data, limiting their relevance for real-world enterprise systems. In this work, we introduce a novel QAE-based framework designed specifically for MTS AD towards enterprise scale. We theoretically develop and experimentally validate the architecture, demonstrating that our QAE achieves performance competitive with neural-network-based autoencoders while requiring fewer trainable parameters. We evaluate our model on datasets that closely reflect SAP system telemetry and show that the proposed QAE is a viable and efficient alternative for semisupervised AD in real-world enterprise settings.

Paper Structure

This paper contains 27 sections, 11 equations, 4 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Contributions
  4. Preliminaries
  5. Anomalies
  6. Point Anomalies
  7. Contextual Anomalies
  8. Collective Anomalies
  9. Multivariate Time Series
  10. Window-Based Anomaly Detection
  11. Autoencoders
  12. Quantum Autoencoder for Multivariate Time Series Anomaly Detection
  13. Quantum Autoencoder Architecture
  14. Model Training and Inference
  15. Experiments
  16. ...and 12 more sections

Figures (4)

  • Figure 1: General architecture of an AE consisting of encoder and decoder with the bottleneck layer.
  • Figure 2: The QAE is realized using a trainable re-upload encoding architecture including multiplicative weight and additive bias parameters. Detailed explanation in text.
  • Figure 3: Violin plots illustrating the distribution of reconstruction errors of QAE on two different machines of the SMD dataset.
  • Figure 4: Reconstruction errors of QAE and the large AE on B1 of the Pasta dataset.