Recently, multivariate time-series data has been generated in various environments, such as sensor networks and IoT, making anomaly detection in time-series data an essential research topic. Unsupervised learning anomaly detectors identify anomalies by training a model on normal data and producing high residuals for abnormal observations. However, a fundamental issue arises as anomalies do not consistently result in high residuals, necessitating a focus on the time-series patterns of residuals rather than individual residual sizes. In this paper, we present a novel framework comprising two serialized anomaly detectors: the first model calculates residuals as usual, while the second one evaluates the time-series pattern of the computed residuals to determine whether they are normal or abnormal. Experiments conducted on real-world time-series data demonstrate the effectiveness of our proposed framework.
Byeongtae PARK
Hanyang University
Dong-Kyu CHAE
Hanyang University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Byeongtae PARK, Dong-Kyu CHAE, "A Novel Anomaly Detection Framework Based on Model Serialization" in IEICE TRANSACTIONS on Information,
vol. E107-D, no. 3, pp. 420-423, March 2024, doi: 10.1587/transinf.2023EDL8024.
Abstract: Recently, multivariate time-series data has been generated in various environments, such as sensor networks and IoT, making anomaly detection in time-series data an essential research topic. Unsupervised learning anomaly detectors identify anomalies by training a model on normal data and producing high residuals for abnormal observations. However, a fundamental issue arises as anomalies do not consistently result in high residuals, necessitating a focus on the time-series patterns of residuals rather than individual residual sizes. In this paper, we present a novel framework comprising two serialized anomaly detectors: the first model calculates residuals as usual, while the second one evaluates the time-series pattern of the computed residuals to determine whether they are normal or abnormal. Experiments conducted on real-world time-series data demonstrate the effectiveness of our proposed framework.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2023EDL8024/_p
Copy
@ARTICLE{e107-d_3_420,
author={Byeongtae PARK, Dong-Kyu CHAE, },
journal={IEICE TRANSACTIONS on Information},
title={A Novel Anomaly Detection Framework Based on Model Serialization},
year={2024},
volume={E107-D},
number={3},
pages={420-423},
abstract={Recently, multivariate time-series data has been generated in various environments, such as sensor networks and IoT, making anomaly detection in time-series data an essential research topic. Unsupervised learning anomaly detectors identify anomalies by training a model on normal data and producing high residuals for abnormal observations. However, a fundamental issue arises as anomalies do not consistently result in high residuals, necessitating a focus on the time-series patterns of residuals rather than individual residual sizes. In this paper, we present a novel framework comprising two serialized anomaly detectors: the first model calculates residuals as usual, while the second one evaluates the time-series pattern of the computed residuals to determine whether they are normal or abnormal. Experiments conducted on real-world time-series data demonstrate the effectiveness of our proposed framework.},
keywords={},
doi={10.1587/transinf.2023EDL8024},
ISSN={1745-1361},
month={March},}
Copy
TY - JOUR
TI - A Novel Anomaly Detection Framework Based on Model Serialization
T2 - IEICE TRANSACTIONS on Information
SP - 420
EP - 423
AU - Byeongtae PARK
AU - Dong-Kyu CHAE
PY - 2024
DO - 10.1587/transinf.2023EDL8024
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E107-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2024
AB - Recently, multivariate time-series data has been generated in various environments, such as sensor networks and IoT, making anomaly detection in time-series data an essential research topic. Unsupervised learning anomaly detectors identify anomalies by training a model on normal data and producing high residuals for abnormal observations. However, a fundamental issue arises as anomalies do not consistently result in high residuals, necessitating a focus on the time-series patterns of residuals rather than individual residual sizes. In this paper, we present a novel framework comprising two serialized anomaly detectors: the first model calculates residuals as usual, while the second one evaluates the time-series pattern of the computed residuals to determine whether they are normal or abnormal. Experiments conducted on real-world time-series data demonstrate the effectiveness of our proposed framework.
ER -