When spectral libraries are too complex to search: Evolutionary subset selection for domain-adaptive calibration
Analytica Chimica Acta · 2026-08 · 1412 · 345651
DOI: 10.1016/j.aca.2026.345651 ↗Kurz gefasst
Fachartikel in der Zeitschrift Analytica Chimica Acta, 2026-08. Verfasst von Leonardo Ramirez-Lopez und 11 weiteren Autor:innen. Beteiligte Einrichtung: Imperial College London, Imperial College Business School, South Kensington Campus, London, SW7 2AZ, England, United Kingdom; BUCHI Labortechnik AG, …. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Nuclear Science and Security Consortium, Natural Resources Conservation Service, Eidgenössische Technische Hochschule Zürich.
Abstract (Originalsprache)
BACKGROUND: The increasing availability of large spectral libraries offers new opportunities to reduce the costs and efforts required to develop spectroscopy-based sensing techniques for rapidly and non-destructively estimating key properties across environmental and agricultural matrices. These libraries can provide training samples for developing models adapted to specific target domains. However, identifying which samples are most relevant for a given target domain remains challenging. This study introduces gesearch, a non-linear evolutionary algorithm for selecting target-domain-relevant training samples from complex spectral libraries to build accurate and interpretable quantitative models. RESULTS: The gesearch, method was used to extract a subset of relevant samples from a large North American infrared soil spectral library to build predictive models of total carbon for an independent target area in the Democratic Republic of the Congo. In this challenging cross-domain test case, simple linear models built with the training samples found by gesearch achieved superior accuracy compared with established modelling approaches, including LOCAL, Cubist, convolutional neural networks, and global partial least squares regression. SIGNIFICANCE: The proposed method provides a practical framework for exploiting large heterogeneous spectral libraries when only limited target-domain reference data are available. By selecting samples that are spectrally and compositionally coherent with the target domain, gesearch supports accurate, compact, and interpretable calibration models. It can also operate when only unlabelled target-domain spectra are available.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Imperial College London, Imperial College Business School, South Kensington Campus, London, SW7 2AZ, England, United Kingdom; BUCHI Labortechnik AG, Department of Data Science, Meierseggstrasse 40, Flawil, CH-9230, Swit…
- Curtin University, Soil and Landscape Science, School of Molecular and Life Sciences, GPO Box U1987, Perth, WA 6845, Australia.
- BUCHI Labortechnik AG, Department of Data Science, Meierseggstrasse 40, Flawil, CH-9230, Switzerland.
- College of Science and Engineering, James Cook University, Cairns, Queensland, Australia. Electronic address: alexandre.wadoux@jcu.edu.au.
- European Commission, Joint Research Centre (JRC), Via E. Fermi, 2749, Ispra, 21027, Italy.
- Research Institute of Organic Agriculture FiBL, Department of Soil Sciences, Ackerstrasse 113, Frick, CH-5070, Switzerland.
Förderung
- Nuclear Science and Security Consortium
- Natural Resources Conservation Service
- Eidgenössische Technische Hochschule Zürich
Zitieren
APA
Ramirez-Lopez, L., Viscarra Rossel, R., Orellano, C., Kooijman, L., Perez-Fernandez, E., Wadoux, A. M. J., Plans, M., Breure, T., Summerauer, L., Safanelli, J. L., Behrens, T., & Boqué, R. (2026). When spectral libraries are too complex to search: Evolutionary subset selection for domain-adaptive calibration. Analytica Chimica Acta, 1412, 345651. https://doi.org/10.1016/j.aca.2026.345651
Vancouver
Ramirez-Lopez L, Viscarra Rossel R, Orellano C, Kooijman L, Perez-Fernandez E, Wadoux AMJ, et al. When spectral libraries are too complex to search: Evolutionary subset selection for domain-adaptive calibration. Analytica Chimica Acta. 2026;1412:345651. doi:10.1016/j.aca.2026.345651
BibTeX
@article{RamirezLopez2026,
author = {Ramirez-Lopez, Leonardo and Viscarra Rossel, Raphael and Orellano, Claudio and Kooijman, Laurens and Perez-Fernandez, Estefania and Wadoux, Alexandre M J-C and Plans, Marcal and Breure, Timo and Summerauer, Laura and Safanelli, José L and Behrens, Thorsten and Boqué, Ricard},
title = {When spectral libraries are too complex to search: Evolutionary subset selection for domain-adaptive calibration},
journal = {Analytica Chimica Acta},
year = {2026},
volume = {1412},
pages = {345651},
doi = {10.1016/j.aca.2026.345651}
}
EUVIMED