Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests
Monitoring of streamed data to detect abnormal behaviour (variously known as event detection, anomaly detection, change detection, or outlier detection) underlies many applications, especially within the Internet of Things. There, one often collects data from a variety of sources, with asynchronous...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9715127/ |
| _version_ | 1811272801833713664 |
|---|---|
| author | Jakub Marecek Stathis Maroulis Vana Kalogeraki Dimitrios Gunopulos |
| author_facet | Jakub Marecek Stathis Maroulis Vana Kalogeraki Dimitrios Gunopulos |
| author_sort | Jakub Marecek |
| collection | DOAJ |
| description | Monitoring of streamed data to detect abnormal behaviour (variously known as event detection, anomaly detection, change detection, or outlier detection) underlies many applications, especially within the Internet of Things. There, one often collects data from a variety of sources, with asynchronous sampling, and missing data. In this setting, one can detect abnormal behavior using low-rank techniques. In particular, we assume that normal observations come from a low-rank subspace, prior to being corrupted by a uniformly distributed noise. Correspondingly, we aim to recover a representation of the subspace, and perform event detection by running point-to-subspace distance query for incoming data. We use a variant of low-rank factorisation, which considers interval uncertainty sets around “known entries”, on a suitable flattening of the input data to obtain a low-rank model. On-line, we compute the distance of incoming data to the low-rank normal subspace and update the subspace to keep it consistent with the seasonal changes present. For the distance computation, we consider subsampling. We bound the one-sided error as a function of the number of coordinates employed. In our computational experiments, we test the proposed algorithm on induction-loop data from Dublin, Ireland. |
| first_indexed | 2024-04-12T22:46:50Z |
| format | Article |
| id | doaj.art-3c1847caa20b487fba2fa1d5eb57d553 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-12T22:46:50Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-3c1847caa20b487fba2fa1d5eb57d5532022-12-22T03:13:30ZengIEEEIEEE Access2169-35362022-01-0110325253253610.1109/ACCESS.2022.31522069715127Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity TestsJakub Marecek0https://orcid.org/0000-0003-0839-0691Stathis Maroulis1https://orcid.org/0000-0002-2872-7821Vana Kalogeraki2Dimitrios Gunopulos3Department of Computer Science, Faculty of Electrical Engineering, Czech Technical University in Prague, Prague, Czech RepublicDepartment of Informatics, Athens University of Economics and Business, Athens, GreeceDepartment of Informatics, Athens University of Economics and Business, Athens, GreeceDepartment of Informatics and Telecommunications, National and Kapodistrian University of Athens, Athens, GreeceMonitoring of streamed data to detect abnormal behaviour (variously known as event detection, anomaly detection, change detection, or outlier detection) underlies many applications, especially within the Internet of Things. There, one often collects data from a variety of sources, with asynchronous sampling, and missing data. In this setting, one can detect abnormal behavior using low-rank techniques. In particular, we assume that normal observations come from a low-rank subspace, prior to being corrupted by a uniformly distributed noise. Correspondingly, we aim to recover a representation of the subspace, and perform event detection by running point-to-subspace distance query for incoming data. We use a variant of low-rank factorisation, which considers interval uncertainty sets around “known entries”, on a suitable flattening of the input data to obtain a low-rank model. On-line, we compute the distance of incoming data to the low-rank normal subspace and update the subspace to keep it consistent with the seasonal changes present. For the distance computation, we consider subsampling. We bound the one-sided error as a function of the number of coordinates employed. In our computational experiments, we test the proposed algorithm on induction-loop data from Dublin, Ireland.https://ieeexplore.ieee.org/document/9715127/Multidimensional signal processingmonitoringmatrix completionpoint-to-subspace proximityprobably approximately correct learning |
| spellingShingle | Jakub Marecek Stathis Maroulis Vana Kalogeraki Dimitrios Gunopulos Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests IEEE Access Multidimensional signal processing monitoring matrix completion point-to-subspace proximity probably approximately correct learning |
| title | Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests |
| title_full | Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests |
| title_fullStr | Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests |
| title_full_unstemmed | Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests |
| title_short | Low-Rank Methods in Event Detection With Subsampled Point-to-Subspace Proximity Tests |
| title_sort | low rank methods in event detection with subsampled point to subspace proximity tests |
| topic | Multidimensional signal processing monitoring matrix completion point-to-subspace proximity probably approximately correct learning |
| url | https://ieeexplore.ieee.org/document/9715127/ |
| work_keys_str_mv | AT jakubmarecek lowrankmethodsineventdetectionwithsubsampledpointtosubspaceproximitytests AT stathismaroulis lowrankmethodsineventdetectionwithsubsampledpointtosubspaceproximitytests AT vanakalogeraki lowrankmethodsineventdetectionwithsubsampledpointtosubspaceproximitytests AT dimitriosgunopulos lowrankmethodsineventdetectionwithsubsampledpointtosubspaceproximitytests |