Embracing imperfection: machine-assisted invertebrate classification in difficult datasets
J. D. Blair, M. D. Weiser, K. deBeurs, M. Kaspari, C. D. Siler, K. E. Marshall
This paper trains machine learning models to classify terrestrial invertebrates using morphometric data and contextual metadata, achieving 72.7% accuracy at the lowest taxonomic level.
Ecological Informatics · Dec 1, 2022Peer-reviewedEcology