How to treat mixed behavior segments in supervised machine learning of behavioural modes from inertial measurement data.
Animal behaviour
Bio-logging
Body-acceleration
Machine learning
Journal
Movement ecology
ISSN: 2051-3933
Titre abrégé: Mov Ecol
Pays: England
ID NLM: 101635009
Informations de publication
Date de publication:
10 Jun 2024
10 Jun 2024
Historique:
received:
21
02
2024
accepted:
06
06
2024
medline:
11
6
2024
pubmed:
11
6
2024
entrez:
10
6
2024
Statut:
epublish
Résumé
The application of supervised machine learning methods to identify behavioural modes from inertial measurements of bio-loggers has become a standard tool in behavioural ecology. Several design choices can affect the accuracy of identifying the behavioural modes. One such choice is the inclusion or exclusion of segments consisting of more than a single behaviour (mixed segments) in the machine learning model training data. Currently, the common practice is to ignore such segments during model training. In this paper we tested the hypothesis that including mixed segments in model training will improve accuracy, as the model would perform better in identifying them in the test data. We test this hypothesis using a series of data simulations on four datasets of accelerometer data coupled with behaviour observations, obtained from four study species (Damaraland mole-rats, meerkats, olive baboons, polar bears). Results show that when a substantial proportion of the test data are mixed behaviour segments (above ~ 10%), including mixed segments in machine learning model training improves the accuracy of classification. These results were consistent across the four study species, and robust to changes in segment length, sample size, and degree of mixture within the mixed segments. However, we also find that in some cases (particularly in baboons) models trained with mixed segments show reduced accuracy in classifying test data containing only single behaviour (pure) segments, compared to models trained without mixed segments. Based on these results, we recommend that when the classification model is expected to deal with a substantial proportion of mixed behaviour segments (> 10%), it is beneficial to include them in model training, otherwise, it is unnecessary but also not harmful. The exception is when there is a basis to assume that the training data contains a higher rate of mixed segments than the actual (unobserved) data to be classified-such a situation may occur particularly when training data are collected in captivity and used to classify data from the wild. In this case, excess inclusion of mixed segments in training data should probably be avoided.
Identifiants
pubmed: 38858733
doi: 10.1186/s40462-024-00485-7
pii: 10.1186/s40462-024-00485-7
doi:
Types de publication
Journal Article
Langues
eng
Pagination
44Informations de copyright
© 2024. The Author(s).
Références
Kays R, Crofoot MC, Jetz W, Wikelski M. Terrestrial animal tracking as an eye on life and planet. Science 2015;348.
Nathan R, Monk CT, Arlinghaus R, Adam T, Alós J, Assaf M, Baktoft H, Beardsworth CE, Bertram MG, Bijleveld AI. Big-data approaches lead to an increased understanding of the ecology of animal movement. Science 2022;375:eabg1780.
Brown DD, Kays R, Wikelski M, Wilson R, Klimley AP. Observing the unwatchable through acceleration logging of animal behavior. Anim Biotelemetry. 2013;1:20.
doi: 10.1186/2050-3385-1-20
Nathan R, Spiegel O, Fortmann-Roe S, Harel R, Wikelski M, Getz WM. Using tri-axial acceleration data to identify behavioral modes of free-ranging animals: general concepts and tools illustrated for griffon vultures. J Exp Biol. 2012;215:986–96.
doi: 10.1242/jeb.058602
pubmed: 22357592
pmcid: 3284320
Resheff YS, Rotics S, Harel R, Spiegel O, Nathan R. AcceleRater: a web application for supervised learning of behavioral modes from acceleration measurements. Mov Ecol. 2014;2:27.
doi: 10.1186/s40462-014-0027-0
pubmed: 25709835
pmcid: 4337760
Yu H, Klaassen M. R package for animal behavior classification from accelerometer data—rabc. Ecol Evol. 2021;11:12364–77.
doi: 10.1002/ece3.7937
pubmed: 34594505
pmcid: 8462134
Resheff YS, Bensch HM, Zöttl M, Rotics S. Correcting a bias in the computation of behavioural time budgets that are based on supervised learning. Methods Ecol Evol. 2022;13:1488–96.
doi: 10.1111/2041-210X.13862
Bom RA, Bouten W, Piersma T, Oosterbeek K, van Gils JA. Optimizing acceleration-based ethograms: the use of variable-time versus fixed-time segmentation. Mov Ecol. 2014;2:6.
doi: 10.1186/2051-3933-2-6
pubmed: 25520816
pmcid: 4267607
Clarke TM, Whitmarsh SK, Hounslow JL, Gleiss AC, Payne NL, Huveneers C. Using tri-axial accelerometer loggers to identify spawning behaviours of large pelagic fish. Movement Ecol. 2021;9.
Hammond TT, Springthorpe D, Walsh RE, Berg-Kirkpatrick T. Using accelerometers to remotely and automatically characterize behavior in small animals. J Exp Biol. 2016;219:1618–24.
pubmed: 26994177
Buitinck L, Louppe G, Blondel M, Pedregosa F, Mueller A, Grisel O, Niculae V, Prettenhofer P, Gramfort A, Grobler J, Layton R. API design for machine learning software: experiences from the scikit-learn project. 2013. arXiv preprint arXiv:1309.0238
Houslay TM, Vullioud P, Zöttl M, Clutton-Brock TH. Benefits of cooperation in captive Damaraland mole-rats. Behav Ecol. 2020;31:711–8.
doi: 10.1093/beheco/araa015
Zöttl M, Vullioud P, Mendonca R, Tico MT, Gaynor D, Mitchell A, Clutton-Brock T. Differences in cooperative behavior among Damaraland mole rats are consequences of an age-related polyethism. Proc Natl Acad Sci USA. 2016;113:10382–7.
doi: 10.1073/pnas.1607885113
pubmed: 27588902
pmcid: 5027444
Friard O, Gamba M. BORIS: a free, versatile open-source event-logging software for video/audio coding and live observations. Methods Ecol Evol. 2016;7:1325–30.
doi: 10.1111/2041-210X.12584
Clutton-Brock T, Manser M. Meerkats: cooperative breeding in the Kalahari. In: Koenig WD, Dickinson JL, editors. Cooperative breeding in vertebrates: studies of ecology, evolution, and behavior. Cambridge: Cambridge University Press; 2016. p. 294–317.
doi: 10.1017/CBO9781107338357.018
Groenewoud F, Clutton-Brock T. Meerkat helpers buffer the detrimental effects of adverse environmental conditions on fecundity, growth and survival. J Anim Ecol. 2021;90:641–52.
doi: 10.1111/1365-2656.13396
pubmed: 33241582
Suire A, Kunita I, Harel R, Crofoot M, Mutinda M, Kamau M, Hassel JM, Murray S, Kawamura S, Matsumoto-Oda A. Estimating individual exposure to predation risk in group-living baboons. Papio anubis Plos One. 2023;18:e0287357.
doi: 10.1371/journal.pone.0287357
pubmed: 37939092
Pagano AM, Rode KD, Cutting A, Owen MA, Jensen S, Ware JV, Robbins C, Durner GM, Atwood TC, Obbard M. Using tri-axial accelerometers to identify wild polar bear behaviors. Endangered Species Res. 2017;32:19–33.
doi: 10.3354/esr00779
Hoffman B, Cusimano M, Baglione V, Canestrari D, Chevallier D, DeSantis DL, Jeantet L, Ladds MA, Maekawa T, Mata-Silva V. A benchmark for computational analysis of animal behavior, using animal-borne tags; 2023. arXiv preprint arXiv:2305.10740
Thiebault A, Huetz C, Pistorius P, Aubin T, Charrier I. Animal-borne acoustic data alone can provide high accuracy classification of activity budgets. Anim Biotelemetry 2021;9.
Weegman MD, Bearhop S, Hilton GM, Walsh AJ, Griffin L, Resheff YS, Nathan R, Fox AD. Using accelerometry to compare costs of extended migration in an arctic herbivore. Curr Zool. 2017;63:667–74.
doi: 10.1093/cz/zox056
pubmed: 29492028
pmcid: 5804222
Chakravarty P, Cozzi G, Ozgul A, Aminian K. A novel biomechanical approach for animal behaviour recognition using accelerometers. Methods Ecol Evol. 2019;10:802–14.
doi: 10.1111/2041-210X.13172
Pagano AM, Atwood TC, Durner GM, Williams TM. The seasonal energetic landscape of an apex marine carnivore, the polar bear. Ecology 2020;101.
Studd EK, Derbyshire RE, Menzies AK, Simms JF, Humphries MM, Murray DL, Boutin S. The Purr-fect Catch: using accelerometers and audio recorders to document kill rates and hunting behaviour of a small prey specialist. Methods Ecol Evol. 2021;12:1277–87.
doi: 10.1111/2041-210X.13605
Dickinson ER, Twining JP, Wilson R, Stephens PA, Westander J, Marks N, Scantlebury DM. Limitations of using surrogates for behaviour classification of accelerometer data: refining methods using random forest models in Caprids. 2021;Movement Ecol. 9.
Hanscom RJ, DeSantis DL, Hill JL, Marbach T, Sukumaran J, Tipton AF, Thompson ML, Higham TE, Clark RW. How to study a predator that only eats a few meals a year: high-frequency accelerometry to quantify feeding behaviours of rattlesnakes (Crotalus spp.). Anim Biotelemetry 2023;11.
Lear KO, Morgan DL, Whitty JM, Whitney NM, Byrnes EE, Beatty SJ, Gleiss AC. Divergent field metabolic rates highlight the challenges of increasing temperatures and energy limitation in aquatic ectotherms. Oecologia. 2020;193:311–23.
doi: 10.1007/s00442-020-04669-x
pubmed: 32435843
Rotics S, Kaatz M, Resheff YS, Turjeman SF, Zurell D, Sapir N, Eggers U, Flack A, Fiedler W, Jeltsch F, Wikelski M, Nathan R. The challenges of the first migration: movement and behaviour of juvenile vs. adult white storks with insights regarding juvenile mortality. J Anim Ecol. 2016;85:938–47.
doi: 10.1111/1365-2656.12525
pubmed: 27046512