A Novel Hybrid Methodology for Anomaly Detection in Time Series
Abstract Numerous research methods have been developed to detect anomalies in the areas of security and risk analysis. In healthcare, there are numerous use cases where anomaly detection is relevant. For example, early detection of sepsis is one such use case. Early treatment of sepsis is cost effec...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2022-07-01
|
| Series: | International Journal of Computational Intelligence Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44196-022-00100-w |
| _version_ | 1828259804009725952 |
|---|---|
| author | Lejla Begic Fazlic Ahmed Halawa Anke Schmeink Robert Lipp Lukas Martin Arne Peine Marlies Morgen Thomas Vollmer Stefan Winter Guido Dartmann |
| author_facet | Lejla Begic Fazlic Ahmed Halawa Anke Schmeink Robert Lipp Lukas Martin Arne Peine Marlies Morgen Thomas Vollmer Stefan Winter Guido Dartmann |
| author_sort | Lejla Begic Fazlic |
| collection | DOAJ |
| description | Abstract Numerous research methods have been developed to detect anomalies in the areas of security and risk analysis. In healthcare, there are numerous use cases where anomaly detection is relevant. For example, early detection of sepsis is one such use case. Early treatment of sepsis is cost effective and reduces the number of hospital days of patients in the ICU. There is no single procedure that is sufficient for sepsis diagnosis, and combinations of approaches are needed. Detecting anomalies in patient time series data could help speed the development of some decisions. However, our algorithm must be viewed as complementary to other approaches based on laboratory values and physician judgments. The focus of this work is to develop a hybrid method for detecting anomalies that occur, for example, in multidimensional medical signals, sensor signals, or other time series in business and nature. The novelty of our approach lies in the extension and combination of existing approaches: Statistics, Self Organizing Maps and Linear Discriminant Analysis in a unique and unprecedented way with the goal of identifying different types of anomalies in real-time measurement data and defining the point where the anomaly occurs. The proposed algorithm not only has the full potential to detect anomalies, but also to find real points where an anomaly starts. |
| first_indexed | 2024-04-13T03:15:22Z |
| format | Article |
| id | doaj.art-821bf39be0b84a968e6ceb6ee4582a63 |
| institution | Directory Open Access Journal |
| issn | 1875-6883 |
| language | English |
| last_indexed | 2024-04-13T03:15:22Z |
| publishDate | 2022-07-01 |
| publisher | Springer |
| record_format | Article |
| series | International Journal of Computational Intelligence Systems |
| spelling | doaj.art-821bf39be0b84a968e6ceb6ee4582a632022-12-22T03:04:56ZengSpringerInternational Journal of Computational Intelligence Systems1875-68832022-07-0115111610.1007/s44196-022-00100-wA Novel Hybrid Methodology for Anomaly Detection in Time SeriesLejla Begic Fazlic0Ahmed Halawa1Anke Schmeink2Robert Lipp3Lukas Martin4Arne Peine5Marlies Morgen6Thomas Vollmer7Stefan Winter8Guido Dartmann9Environmental Campus Birkenfeld, Trier University of Applied SciencesISEK Research Area, RWTH Aachen UniversityISEK Research Area, RWTH Aachen UniversityISEK Research Area, RWTH Aachen UniversityDepartment of Intensive and Intermediate Care, University Hospital AachenDepartment of Intensive and Intermediate Care, University Hospital AachenEnvironmental Campus Birkenfeld, Trier University of Applied SciencesPhilips GmbH Innovative TechnologiesPhilips GmbH Innovative TechnologiesEnvironmental Campus Birkenfeld, Trier University of Applied SciencesAbstract Numerous research methods have been developed to detect anomalies in the areas of security and risk analysis. In healthcare, there are numerous use cases where anomaly detection is relevant. For example, early detection of sepsis is one such use case. Early treatment of sepsis is cost effective and reduces the number of hospital days of patients in the ICU. There is no single procedure that is sufficient for sepsis diagnosis, and combinations of approaches are needed. Detecting anomalies in patient time series data could help speed the development of some decisions. However, our algorithm must be viewed as complementary to other approaches based on laboratory values and physician judgments. The focus of this work is to develop a hybrid method for detecting anomalies that occur, for example, in multidimensional medical signals, sensor signals, or other time series in business and nature. The novelty of our approach lies in the extension and combination of existing approaches: Statistics, Self Organizing Maps and Linear Discriminant Analysis in a unique and unprecedented way with the goal of identifying different types of anomalies in real-time measurement data and defining the point where the anomaly occurs. The proposed algorithm not only has the full potential to detect anomalies, but also to find real points where an anomaly starts.https://doi.org/10.1007/s44196-022-00100-wAnomaly detectionClassificationSelf Organizing Maps (SOM)Linear Discriminant Analysis (LDA) |
| spellingShingle | Lejla Begic Fazlic Ahmed Halawa Anke Schmeink Robert Lipp Lukas Martin Arne Peine Marlies Morgen Thomas Vollmer Stefan Winter Guido Dartmann A Novel Hybrid Methodology for Anomaly Detection in Time Series International Journal of Computational Intelligence Systems Anomaly detection Classification Self Organizing Maps (SOM) Linear Discriminant Analysis (LDA) |
| title | A Novel Hybrid Methodology for Anomaly Detection in Time Series |
| title_full | A Novel Hybrid Methodology for Anomaly Detection in Time Series |
| title_fullStr | A Novel Hybrid Methodology for Anomaly Detection in Time Series |
| title_full_unstemmed | A Novel Hybrid Methodology for Anomaly Detection in Time Series |
| title_short | A Novel Hybrid Methodology for Anomaly Detection in Time Series |
| title_sort | novel hybrid methodology for anomaly detection in time series |
| topic | Anomaly detection Classification Self Organizing Maps (SOM) Linear Discriminant Analysis (LDA) |
| url | https://doi.org/10.1007/s44196-022-00100-w |
| work_keys_str_mv | AT lejlabegicfazlic anovelhybridmethodologyforanomalydetectionintimeseries AT ahmedhalawa anovelhybridmethodologyforanomalydetectionintimeseries AT ankeschmeink anovelhybridmethodologyforanomalydetectionintimeseries AT robertlipp anovelhybridmethodologyforanomalydetectionintimeseries AT lukasmartin anovelhybridmethodologyforanomalydetectionintimeseries AT arnepeine anovelhybridmethodologyforanomalydetectionintimeseries AT marliesmorgen anovelhybridmethodologyforanomalydetectionintimeseries AT thomasvollmer anovelhybridmethodologyforanomalydetectionintimeseries AT stefanwinter anovelhybridmethodologyforanomalydetectionintimeseries AT guidodartmann anovelhybridmethodologyforanomalydetectionintimeseries AT lejlabegicfazlic novelhybridmethodologyforanomalydetectionintimeseries AT ahmedhalawa novelhybridmethodologyforanomalydetectionintimeseries AT ankeschmeink novelhybridmethodologyforanomalydetectionintimeseries AT robertlipp novelhybridmethodologyforanomalydetectionintimeseries AT lukasmartin novelhybridmethodologyforanomalydetectionintimeseries AT arnepeine novelhybridmethodologyforanomalydetectionintimeseries AT marliesmorgen novelhybridmethodologyforanomalydetectionintimeseries AT thomasvollmer novelhybridmethodologyforanomalydetectionintimeseries AT stefanwinter novelhybridmethodologyforanomalydetectionintimeseries AT guidodartmann novelhybridmethodologyforanomalydetectionintimeseries |