EUVIMEDObjevte EUVIMED
Open access · CC BY 4.0

Supervised versus unsupervised approaches to classification of accelerometry data

Maitreyi Sur, Jonathan C Hall, Joseph Brandt, Molly Astell, Sharon A Poessel iD, Todd E Katzner iD

Ecology and Evolution · 2023-05 · 13 (5)

DOI: 10.1002/ece3.10035 ↗

At a glance

Journal article in Ecology and Evolution, published online on 2023-05-17. Written by Maitreyi Sur and 5 co-authors. Institution: Conservation Science Global, Inc. West Cape May New Jersey USA. Cited 17 times so far (Crossref); 60 references. The full text is freely available (open access, CC BY 4.0). Funded by: California Department of Fish and Wildlife, National Fish and Wildlife Foundation.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Sophisticated animal‐borne sensor systems are increasingly providing novel insight into how animals behave and move. Despite their widespread use in ecology, the diversity and expanding quality and quantity of data they produce have created a need for robust analytical methods for biological interpretation. Machine learning tools are often used to meet this need. However, their relative effectiveness is not well known and, in the case of unsupervised tools, given that they do not use validation data, their accuracy can be difficult to assess. We evaluated the effectiveness of supervised ( n = 6), semi‐supervised ( n = 1), and unsupervised ( n = 2) approaches to analyzing accelerometry data collected from critically endangered California condors ( Gymnogyps californianus ). Unsupervised K‐means and EM (expectation–maximization) clustering approaches performed poorly, with adequate classification accuracies of 0.81. Kappa statistics were also highest for RF and kNN, in most cases substantially greater than for other modeling approaches. Unsupervised modeling, which is commonly used for the classification of a priori‐defined behaviors in telemetry data, can provide useful information but likely is instead better suited to post hoc definition of generalized behavioral states. This work also shows the potential for substantial variation in classification accuracy among different machine learning approaches and among different metrics of accuracy. As such, when analyzing biotelemetry data, best practices appear to call for the evaluation of several machine learning techniques and several measures of accuracy for each dataset under consideration.

Subjects

Author keywords

California condoraccelerometeranimal behaviorclassificationmachine learning

Authors and institutions

  • Conservation Science Global, Inc. West Cape May New Jersey USA.
  • Present address: Radboud Institute for Biological and Environmental Sciences (RIBES) Radboud University Nijmegen The Netherlands.
  • Department of Biology Eastern Michigan University Ypsilanti Michigan USA.
  • U.S. Fish and Wildlife Service, Hopper Mountain National Wildlife Refuge Complex Ventura California USA.
  • Department of Biology Boise State University Boise Idaho USA.
  • U.S. Geological Survey, Forest and Rangeland Ecosystem Science Center Boise Idaho USA.

ORCID profiles: Sharon A Poessel, Todd E Katzner

Funding

  • California Department of Fish and Wildlife
  • National Fish and Wildlife Foundation

Cite

APA

Sur, M., Hall, J. C., Brandt, J., Astell, M., Poessel, S. A., & Katzner, T. E. (2023). Supervised versus unsupervised approaches to classification of accelerometry data. Ecology and Evolution, 13(5). https://doi.org/10.1002/ece3.10035

Vancouver

Sur M, Hall JC, Brandt J, Astell M, Poessel SA, Katzner TE. Supervised versus unsupervised approaches to classification of accelerometry data. Ecology and Evolution. 2023;13(5). doi:10.1002/ece3.10035

BibTeX
@article{Sur2023,
  author = {Sur, Maitreyi and Hall, Jonathan C and Brandt, Joseph and Astell, Molly and Poessel, Sharon A and Katzner, Todd E},
  title = {Supervised versus unsupervised approaches to classification of accelerometry data},
  journal = {Ecology and Evolution},
  year = {2023},
  volume = {13},
  number = {5},
  doi = {10.1002/ece3.10035}
}
Start for free →