Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • harvard-cite-them-right
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Pattern Detection in Abnormal District Heating Data
University of Borås, Faculty of Librarianship, Information, Education and IT. (CSL@B)ORCID iD: 0000-0002-9685-7775
University of Borås, Faculty of Librarianship, Information, Education and IT. (CSL@B)ORCID iD: 0000-0003-4308-434X
2025 (English)Conference paper, Published paper (Refereed)
Abstract [en]

Data generated by industrial systems such as District Heating (DH) often lack meaningful labels for supervised Machine Learning (ML) methods in anomaly detection. Consequently, unsupervised and semi-supervised ML methods are widely used. These methods frequently uncover numerous anomalies, necessitating labor-intensive post-processing. This paper proposes an algorithm to detect topK anomaly instances with similar patterns (energy signatures) to known anomalies, and to identify clusters of similar anomalies using hierarchical clustering. Similarities between anomaly instances are computed using Dynamic Time Warping and Matrix Profiles. Generative Adversarial Networks (GANs) are employed to augment small anomaly datasets. Results demonstrate the effectiveness of the proposed algorithm in reducing the manual effort required for post-processing anomalies in a DH dataset.

Place, publisher, year, edition, pages
2025. p. 224-239
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:hb:diva-33427DOI: 10.1007/978-3-031-82481-4_16ISI: 001530948600016OAI: oai:DiVA.org:hb-33427DiVA, id: diva2:1951416
Conference
10th International Conference on Machine Learning, Optimization, and Data Science, LOD 2024, Castiglione della Pescaia, 22-25 September 2024
Available from: 2025-04-10 Created: 2025-04-10 Last updated: 2025-10-17Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full text

Authority records

Mbiydzenyuy, GideonSundell, Håkan

Search in DiVA

By author/editor
Mbiydzenyuy, GideonSundell, Håkan
By organisation
Faculty of Librarianship, Information, Education and IT
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 207 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • harvard-cite-them-right
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf