Journal article
Refining Accelerometer‐Based Animal Behaviour Classifications With Sequence‐Informed Post‐Processing
Ecology and Evolution, Vol.16(8), pp.1-16
2026
PMCID: PMC13429299
PMID: 42544294
Appears in UniSC Supported Open Access Outputs
Abstract
Supervised machine learning has been used to detect fine‐scale behaviours from animal‐borne accelerometers by dividing the continuous sequences of behaviour into discrete segments and classifying each with a distinct behavioural category. This approach, while widely implemented, discards the sequential information available in the temporal ordering of the behavioural series. ‘Post‐processing’ (smoothing and error correction made after the initial classifications) can be used to improve the accuracy of the original predictions by learning from the natural transitions and durations of behaviours. While broadly implemented across other classification domains, this technique has been underutilised for accelerometer‐based animal behaviour classification. In this paper, we compare the performance of five different post‐processors (modal, duration‐based, transition‐based, Hidden Markov Model, and a Naive Bayes smoother) against the original predictions from the base classifier across 15 animal accelerometer datasets across 13 species. Overall, while there was no single post‐processing method that was optimal in every dataset, we find Bayesian smoothing to have the best overall performance improvement (average 6.4% increase in F 1‐score), resulting in ecological interpretation closest to the true data. Requiring no additional data and very little additional computational effort, this preliminary research suggests promising potential for the widespread and accessible inclusion of post‐processing in the animal behaviour classification pipeline.
Details
- Title
- Refining Accelerometer‐Based Animal Behaviour Classifications With Sequence‐Informed Post‐Processing
- Authors
- Oakleigh Wilson (Corresponding Author) - University of the Sunshine CoastHui Yu - Deakin UniversityDavid Schoeman - University of the Sunshine CoastGabriella Sparkes - The University of QueenslandChristofer Clemente - University of the Sunshine Coast
- Publication details
- Ecology and Evolution, Vol.16(8), pp.1-16
- Publisher
- John Wiley & Sons Ltd.
- Date published
- 2026
- DOI
- 10.1002/ece3.74053
- ISSN
- 2045-7758
- PMID
- 42544294; PMC13429299
- Copyright note
- © 2026 The Author(s). Ecology and Evolution published by British Ecological Society and John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
- Data Availability
- All raw data was sourced from open-source databases associated with the publications referenced in this manuscript. All code is available at https://github.com/OakAlice/PostProcessing. A single dataset example has been made available on the GitHub.
- Grant note
- This work was supported by Ecological Society of Australia Incorporated (Student Research Grant, Ecological Society of Australia).
- Organisation Unit
- School of Science, Technology and Engineering; Centre for Bioinnovation
- Language
- English
- Record Identifier
- 991250416202621
- Output Type
- Journal article
Metrics
1 Record Views