Søk i bibliotekbasen til NVE
  • Portal
  • Søk
  • Mine lån
  • Utlån
  • Innlevering
  • Gi vurdering
  • Reserver
  • Fjernlån
  • Marcdata
  • RIS
Tittel:Review of methods for automation of quality control on hydrologic time series and considerations for a research approach at NVE
Ansvar:author: Trond Reitan
Forfatter:Reitan, Trond
Materialtype:NVEs utgivelser
Utgitt:Oslo : NVE, 2022
Omfang:46 s. - ill. (noen kol.)
ISBN/ISSN:978-82-410-2206-7 / 1501-2832
Serie:NVE Rapport ; 2022:15
Emneord:Analyser / Anomaly detection / Hydrologi / Hydrological time series / Machine learning / Tidsserier
Geografiske emneord:Norge
Note:Har bibliografi
Innhold:Sammendrag: This report contains a literature study on the topic of automatic
anomaly detection in time series using machine learning. The study
has been carried out in order to see what the state of field is, both in academic circles and in real world applications in institutions that gather hydrologic or meteorologic time series. The academic research is far more advanced of what hydrologic and meteorologic institutions are currently implementing, however some initial forays into applying the theory by such institutions have started. The
research is however applied to time series across a much wider spectrum of applications, so it is hard to judge how well a method works for NVE purposes. Therefore, testing at NVE is required to evaluate how well various methods work. Performance should be based on how good the end results are as well as their speed and robustness
Eksemplar:2. (2 tilgjengelige)