Pith. sign in

REVIEW 4 major objections 6 minor 50 references

EUNIS Habitat Maps: Enhancing Thematic and Spatial Resolution for Europe through Machine Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Machine learning maps 260 European habitats at 100-meter resolution across Europe.

desk verdict A genuinely useful continental habitat map product, but the 'independent' validation is partly a hold-out from the same database family and the abstract overstates accuracy. read the letter →

arxiv 2506.13649 v1 pith:TSFTJ3UW submitted 2025-06-16 stat.AP cs.LGphysics.geo-phq-bio.QM

Sara Si-Moussi , Stephan Hennekens , Sander Mücher , Wanda De Keersmaecker , Milan Chytrý , Emiliano Agrillo , Fabio Attorre , Idoia Biurrun
show 12 more authors
Gianmaria Bonari Andraž Čarni Renata Ćušterevska Tetiana Dziuba Klaus Ecker Behlül Güler Ute Jandt Borja Jiménez-Alfaro Jonathan Lenoir Jens-Christian Svenning Grzegorz Swacha Wilfried Thuiller
This is my paper · ORCID
classification stat.APcs.LGphysics.geo-phq-bio.QM
keywords EUNIShabitatshabitatmappingmachinelearningensemblemodellingremotesensingvegetationplotsspatialcross-validationconservationplanning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that Europe can be mapped wall-to-wall for 260 EUNIS habitat types at hierarchical level 3 using machine learning, producing a product that supports conservation policy and the Nature Restoration Law. The authors train an ensemble of tree-based and neural-network classifiers on nearly 600,000 georeferenced vegetation plots whose species compositions were assigned to EUNIS level 3 classes by an expert system, using climate, topography, soil, and satellite-derived predictors at 100-meter resolution. They validate the resulting maps against independent plot datasets from the Netherlands, France, and Austria and against a spatial-block cross-validation, reporting good but uneven performance with different recall-precision trade-offs by habitat formation. If the maps hold, planners gain a spatially explicit, uncertainty-aware baseline for where habitats occur and where they do not, something that has been missing at continental scale.

What carries the argument

The load-bearing mechanism is a per-formation ensemble of multi-class classifiers: for each of the terrestrial EUNIS level 1 groups (saltmarshes, coastal, wetlands, grasslands, shrublands and scrub and tundra, forests, sparsely vegetated, and man-made), a separate model predicts level 3 classes from environmental and satellite predictors, and the models from different algorithm families and spatial cross-validation folds are combined by weighted voting. Two rule-based filters make the continuous probabilities into a map: regional masks based on ecoregions and coastline distance keep predictions inside each class's known range, and a land-cover crosswalk with priority rules decides the prevailing level 1 formation and the final level 3 class at each pixel. Uncertainty is quantified at each pixel as the committee-averaging score and as model- and fold-induced disagreement.

What would settle it

Score the published map against a national habitat map from a country not used in validation (for example, Spain or Denmark) at 100 m resolution; if common habitat classes there score near-zero F1 despite good cross-validation, the continental-scale claim would be falsified. A cheaper check is to count classes with F1 below 0.3 in the provided Austrian 10 m validation, since matching a finer national reference is the hardest test the product faces.

Watch

Extended reading notes

Core claim

The central claim is that a wall-to-wall map of European terrestrial habitats at EUNIS level 3 is attainable by training separate multi-class machine-learning ensembles within each EUNIS level 1 formation, with roughly 597,819 vegetation plots as ground truth once their species lists are translated to level 3 by the EUNIS-ESy expert system. Instead of fitting one binary model per habitat, the authors jointly model the classes in each formation so less common habitats can borrow information from more common ones, then combine tree-based, boosting, and neural-network classifiers using weighted voting over spatial cross-validation folds. The resulting 100 m maps carry per-pixel probabilities, top-3 classes, and confidence scores, and the validation results show high F1 scores for saltmarshes, sparsely vegetated, coastal, and wetland habitats, with grasslands, shrublands, and forests more variable and with different recall-precision trade-offs on independent data from the Netherlands, France, and Austria.

Load-bearing premise

The load-bearing premise is that the expert-system assignments used to train and validate the models are correct enough to serve as ground truth; if those labels are wrong, high validation scores would not transfer to the real world.

Editorial extensions

If this is right

  • A user can download continuous probability layers for all 260 classes and use the top-3 list to identify pixels where the first choice is fragile, making the map usable for targeting field surveys.
  • Conservation and restoration planners can overlay the confidence maps to distinguish well-predicted habitats from uncertain edge cases, instead of treating the single most probable class as truth.
  • Because the mapping workflow separates probabilities from rule-based filtering, substituting a finer land-cover layer than the one used here would yield a finer final product without retraining the models.
  • The per-formation ensembles make it possible to update or add habitat classes within one formation without retraining the whole continent.
  • For formations with few classes or strong abiotic control, per-class F1 scores above 0.9 are common, so the map is already usable for saltmarsh and sparsely vegetated habitats.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable extension not in the paper would be to evaluate the full pipeline on pixels where a secondary formation is nearly as probable as the winner, since errors in choosing the level 1 formation would then propagate into the final level 3 label.
  • The three validation territories cover Atlantic, Alpine, and parts of Mediterranean and Continental Europe; a decisive next test is Scandinavia, Iberia, or Eastern Europe, where plot densities and habitat combinations differ.
  • Adding the missing predictors the authors flag, such as soil moisture, land-use history, and human footprint, and then re-scoring the low-performing humid-soil and abandoned-land classes would show whether the remaining errors are data gaps or model limitations.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This manuscript presents a European wall-to-wall EUNIS habitat map at level 3 with 100 m resolution, produced by training formation-specific multi-class ensemble models on vegetation plots from the European Vegetation Archive (EVA) labeled by the EUNIS-ESy expert system, using climatic, topographic, edaphic, and remote-sensing predictors. The authors report spatial block cross-validation on EVA and external validation against Dutch (NLPT), Austrian (AT), and French forest (IFN) datasets, and they release probability, confidence, and top-habitat map products. The central claims are that the maps achieve strong predictive performance and that the validation includes genuinely independent datasets.

Significance. If the validation claims are supported, the product would be a valuable conservation and restoration planning resource: a continental, high-resolution, wall-to-wall EUNIS level 3 map with uncertainty layers, built from an unusually large training set and a transparent ensemble workflow. The paper also has concrete methodological strengths: spatial block cross-validation is used to mitigate spatial autocorrelation, multiple algorithm families are ensembled, class-imbalance corrections are compared per formation, and the code and data products are publicly available. However, the significance currently rests on two assertions that the evidence in the manuscript does not fully support: the independence of the Dutch validation set and the characterization of predictive performance as 'strong'. The class-level F1 scores reported in Table 7 are moderate for the most extensive formations, and external scores are considerably lower than the cross-validation scores.

major comments (4)
  1. [Habitat datasets for validation] The NLPT dataset is described as an 'independent habitat occurrence dataset' and as a 'hold-out' from the Dutch Landelijke Vegetatie Database (LVD), but reference 29 (Schaminée et al. 2012) is a national vegetation database that is used as a source for EVA, which is the training database for the models. The manuscript does not state that the NLPT plots were removed from EVA before training, nor does it document any independent field collection or labeling procedure for NLPT. Calling this set 'independent' is therefore not justified, and the abstract's claim of 'independent validation' is only supported by the Austrian map and the French IFN forest data. Since NLPT is the only external validation covering non-forest habitats in the Atlantic region, this issue directly affects the credibility of the reported grassland, wetland, and scrub F1 scores.
  2. [Technical Validation, Table 7] The abstract and Technical Validation state that 'the habitat maps obtained strong predictive performances on the validation datasets', but the class-level F1 scores in Table 7 do not support that wording for the dominant formations. Mean forest F1 is 0.61 under EVA cross-validation but only 0.38 (NLPT), 0.33 (AT), and 0.48 (IFN); grassland F1 is 0.66 for EVA but 0.40 (NLPT) and 0.37 (AT); scrub and tundra F1 drops from 0.83 (EVA) to 0.46 (NLPT) and 0.31 (AT). These scores are modest, especially on the independent sets, and the large standard deviations (e.g., 0.32–0.39 for forests) indicate that many individual classes perform poorly. The manuscript should be revised to describe the results as moderate and to highlight the classes or formations where performance is genuinely strong, rather than claiming uniformly strong performance.
  3. [Step 2: Regional filtering rules] The regional masks used in Step 2 are computed from the same EVA vegetation plot occurrences that are also used for cross-validation, but the manuscript does not state whether the EVA cross-validation scores in Tables 7–14 are computed before or after applying these masks. If the masks are applied during validation, then each validation plot's own occurrence contributes to the mask for its ecoregion, which makes the EVA F1 scores optimistic relative to a fully forward prediction. Please clarify the validation protocol and, if masking is included, quantify the effect of the regional masks on the EVA and external validation scores.
  4. [Technical Validation, final paragraph] The statement that low performance 'improved when considering the top three predictions rather than only the most likely class' is not supported by any quantitative comparison in the manuscript. No top-1 versus top-3 accuracy, precision, or F1 table is provided, nor is the improvement quantified for the classes or formations where it is claimed. Since this statement is used to mitigate the low F1 scores, the authors should add a table or figure comparing top-1 and top-3 validation results for at least the independent datasets.
minor comments (6)
  1. [Methods, Ensemble model training] The sentence 'Decision trees excel with structured tabular data and neural networks with intricate feature interactions...' appears twice, once at the end of the 'Ensemble model training' subsection and once immediately before 'Ensemble forecasting and uncertainty'; the duplicate should be removed.
  2. [References, LDAM loss] The citation for LDAM loss (reference 47) points to Cao, Larsen, and Thorne (2001), a paper on rare species in multivariate analysis, which appears to be the wrong reference; the LDAM loss is from Cao et al. (2019), 'Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss' (NeurIPS).
  3. [Table 7] Table 7 contains a stray 'Strategy' row in the header and does not report results for the man-made (V) formation, although Table 1 and the mapping workflow include V; please check the table formatting and clarify whether V results are omitted because no validation data were available.
  4. [Methods, Spatial Block-CV partitioning] The text says the domain is divided into 100 km x 100 km cells that are partitioned into five blocks, and later says '20% of the observations' are hidden per iteration; with five blocks the hidden fraction is 20%, so this is consistent, but the wording should be made explicit to avoid confusion.
  5. [Environmental predictors] Table 2 lists predictors at 1 km, 500 m, and 100 m resolutions, but the manuscript does not describe how these were resampled or harmonized to the 100 m prediction grid; please add a short paragraph on resampling, reprojection, and temporal matching of the predictor layers.
  6. [Tables 1 and 6] Table 6 lists 'P' (inland waters) among the formations for which the same MLP configuration was selected, but Table 1 and the Methods exclude inland waters; this is likely a typo for 'V' and should be corrected.

Circularity Check

1 steps flagged · score 2.0 of 10

Central map derivation is self-contained; the NLPT 'independent' validation is a within-pool hold-out and slightly compromises the validation claim.

  1. other [Methods: 'Habitat datasets for validation'; Abstract; Methods: 'Habitat mapping workflow' Step 2]
    "To evaluate the quality of the habitat maps, we have used two independent habitat occurrence datasets: a hold-out of habitat observations from the Netherlands (NLPT) and the French Forest Inventory (IFN). The NLPT dataset contains 50k vegetation plots (Figure 2) from the Landelijke Vegetatie Database (LVD)29,30"

    The NLPT 'hold-out' comes from LVD, the Dutch national vegetation database that feeds the EVA archive used for training ('Vegetation plots stored in the European Vegetation Archive (EVA) served as ground truth data for training and testing the model across Europe'). A hold-out from the same database pool is a train/test split, not independent validation, so the abstract's 'independent validation' claim is partly circular: the NLPT F1 scores measure within-pool performance. Additionally, Step 2 derives regional masks from 'vegetation plot data from the EVA'; unless NLPT was excluded from that mask computation, the masks used to constrain predictions can encode the NLPT validation labels themselves.

full rationale

The core derivation—EUNIS-ESy-labelled EVA plots combined with environmental and remote-sensing predictors in ensemble classifiers—does not reduce to its inputs by construction: the targets are species-composition assignments from an external expert system, the predictors are independent satellite, climate, topographic, and soil layers, and cross-validation uses spatial blocks. No fitted parameter is renamed as a prediction, and no uniqueness theorem is imported from the authors' prior work. The one circularity-adjacent issue is the NLPT validation set: it is described as 'independent' but is a hold-out from LVD, a component database of EVA used for training; therefore the Dutch F1 scores are a within-pool test, and if the EVA-derived regional masks were built before excluding NLPT, those masks can encode the validation labels. Because AT and IFN remain genuinely external and the map generation itself is not circular, this is a minor validation-independence caveat rather than a collapse of the derivation chain.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The map product rests on expert-system labels, EVA sampling representativeness, and the assumption that environmental and satellite predictors capture habitat-defining variation. The main fitted choices are ML hyperparameters, the 100 km cross-validation block size, the 5-plot occurrence cutoff, and the data-derived regional masks and land cover crosswalks. No new physical or conceptual entities are postulated.

free parameters (6)
  • Focal loss gamma = gamma = 5.0
    Hand-set loss parameter chosen during hyperparameter tuning for neural networks in shrubland and forest; controls how strongly easy examples are down-weighted, so it influences which classes are predicted.
  • LDAM margin = margin = 0.5
    Margin parameter in LDAM loss selected during tuning for several formations; affects class boundary separation for imbalanced classes.
  • Spatial block size = 100 km x 100 km
    Chosen block size for spatial cross-validation. It defines the scale at which spatial autocorrelation is controlled and affects the number of folds and the ensemble diversity.
  • Minimum occurrence threshold = 5 plots
    Classes with fewer than 5 EVA occurrences were discarded, changing the set of mapped classes and making the '260 EUNIS types' claim dependent on this cutoff.
  • Ecoregion association matrix = binary regional masks per class computed from EVA plot occurrences
    Used in Step 2 to filter class probabilities. Because it is derived from the same EVA data used to train the models, it can constrain predictions to known ranges and inflate cross-validated scores.
  • Corine-to-EUNIS crosswalk rules = expert-defined worksheet (output product 4)
    Expert judgment used in Steps 3 and 4 to filter by land cover and to choose the prevailing EUNIS formation; these rules control the final wall-to-wall map.
assumptions (6)
  • domain assumption EUNIS-ESy expert system (Chytrý et al. 2021) correctly assigns EVA vegetation plots to EUNIS level 3 classes.
    All training and validation labels come from this expert system; no independent field verification of labels is provided in this paper.
  • domain assumption The EVA plot set is representative of the diversity and distribution of European terrestrial habitats.
    Models can only learn classes and environments that are sampled in EVA; unsampled regions and rare habitats are missing.
  • domain assumption Climate, topography, soil, hydrography and remote sensing predictors are sufficient to discriminate EUNIS level 3 habitats.
    EUNIS classes are defined partly by species composition, which may not be visible in 100m spectral and environmental data; this limits the achievable accuracy.
  • domain assumption Spatial block CV with 100 km blocks removes bias from spatial autocorrelation.
    The paper assumes 100 km blocks are large enough to prevent data leakage; smaller or larger blocks could change performance estimates.
  • domain assumption The Corine Land Cover crosswalk and priority rules correctly assign each pixel to one EUNIS formation.
    The final wall-to-wall map depends on expert crosswalks between land cover classes and habitats; errors here propagate to the final map.
  • domain assumption Ecological regional masks based on ecoregions and coastline are valid constraints for EUNIS class occurrence.
    The masks restrict predictions to regions where each class was observed; if EVA sampling is incomplete, true occurrences outside masks are impossible to predict.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EUNIS Habitat Maps: Enhancing Thematic and Spatial Resolution for Europe through Machine Learning." pith.science (2026). https://pith.science/paper/TSFTJ3UW

@misc{pith2026250613649,
  author       = {Pith},
  title        = {Pith review of: EUNIS Habitat Maps: Enhancing Thematic and Spatial Resolution for Europe through Machine Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TSFTJ3UW}},
  note         = {Machine review of arXiv:2506.13649}
}
read the original abstract

The EUNIS habitat classification is crucial for categorising European habitats, supporting European policy on nature conservation and implementing the Nature Restoration Law. To meet the growing demand for detailed and accurate habitat information, we provide spatial predictions for 260 EUNIS habitat types at hierarchical level 3, together with independent validation and uncertainty analyses. Using ensemble machine learning models, together with high-resolution satellite imagery and ecologically meaningful climatic, topographic and edaphic variables, we produced a European habitat map indicating the most probable EUNIS habitat at 100-m resolution across Europe. Additionally, we provide information on prediction uncertainty and the most probable habitats at level 3 within each EUNIS level 1 formation. This product is particularly useful for both conservation and restoration purposes. Predictions were cross-validated at European scale using a spatial block cross-validation and evaluated against independent data from France (forests only), the Netherlands and Austria. The habitat maps obtained strong predictive performances on the validation datasets with distinct trade-offs in terms of recall and precision across habitat formations.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 44 canonical work pages

  1. [1]

    The European Environment: State and Outlook 2020 : Knowledge for Transition to a Sustainable Europe

    AEE. The European Environment: State and Outlook 2020 : Knowledge for Transition to a Sustainable Europe. (Publications Office, LU, 2019)

    work page 2020

  2. [2]

    & Davies, C

    Moss, D. & Davies, C. E. EUROPEAN TOPIC CENTRE ON NATURE PROTECTION AND BIODIVERSITY EUNIS HABITAT CLASSIFICATION 2001 WORK PROGRAMME

    work page 2001

  3. [3]

    E., Mo ss, D

    Davies, C. E., Mo ss, D. & Hill, M. O. EUNIS HABITAT CLASSIFICATION REVISED 2004

    work page 2004

  4. [4]

    Chytrý, M. et al. EUNIS Habitat Classification: Expert system, characteristic species combinations and distribution maps of European habitats. Appl. Veg. Sci. 23, 648–675 (2020)

    work page 2020

  5. [5]

    Directorate General for the Environment

    European Commission. Directorate General for the Environment. European Red List of Habitats. Part 2, Terrestrial and Freshwater Habitats. (Publications Office, LU, 2016)

    work page 2016

  6. [6]

    Bohn, U. et al. Karte Der Natürlichen Vegetation Europas / Map of the Natural Vegetation of Europe. (Landwirtschaftsverlag, Münster, 2003)

    work page 2003

  7. [7]

    & Nakhutsrishvili, G

    Bohn, U., Zazanashvili, N. & Nakhutsrishvili, G. The Map of the Natural Vegetation of Europe and its application in the Caucasus Ecoregion. 175, (2007)

    work page 2007

  8. [8]

    Mucina, L. et al. Vegetation of Europe: hierarchical floristic classification system of vascular plant, bryophyte, lichen, and algal communities. Appl. Veg. Sci. 19, 3–264 (2016)

    work page 2016

Show all 50 references
  1. [9]

    Chytrý, M. et al. FloraVeg.EU— An online database of European vegetation, habitats and flora. Appl. Veg. Sci. 27, e12798 (2024)

    2024

  2. [10]

    Chytrý, M. et al. European Vegetation Archive (EVA): an integrated database of European vegetation plots. Appl. Veg. Sci. 19, 173–180 (2016)

    2016

  3. [11]

    Schaminée, J. H. J., Hennekens, S. M. & Ozinga, W. A. Use of the ecological information s ystem SynBioSys for the analysis of large datasets. J. Veg. Sci. 18, 463 –470 (2007)

    2007

  4. [12]

    & Museum national d’Histoire naturelle (MNHN)

    European Environment Agency. & Museum national d’Histoire naturelle (MNHN). Terrestrial Habitat Mapping in Europe: An Overview. (Publications Office, LU, 2014)

    2014

  5. [13]

    & Gamon, J

    Wang, R. & Gamon, J. A. Remote sensing of terrestrial plant biodiversity. Remote Sens. Environ. 231, 111218 (2019)

    2019

  6. [14]

    Adamo, M. et al. Knowledge-Based Classification of Grassland Ecosystem Based on Multi-Temporal WorldView -2 Data and FAO -LCCS Taxonomy. Remote Sens. 12, 1447 (2020)

    2020

  7. [15]

    Wicaksono, P., Aryaguna, P. A. & Lazuardi, W. Benthic Habitat Mapping Model and Cross Validation Using Machine-Learning Classification Algorithms. Remote Sens. 11, 1279 (2019)

    2019

  8. [16]

    Agrillo, E. et al. Earth Observation and Biodiversity Big Data for Forest Habitat Types Classification and Mapping. Remote Sens. 13, 1231 (2021)

    2021

  9. [17]

    Gavish, Y. et al. Comparing the performance of flat and hierarchical Habitat/Land - Cover classification models in a NATURA 2000 site. ISPRS J. Photog ramm. Remote Sens. 136, 1–12 (2018)

    2018

  10. [18]

    & Launeau, P

    Le Dez, M., Robin, M. & Launeau, P. Contribution of Sentinel -2 satellite images for habitat mapping of the Natura 2000 site ‘Estuaire de la Loire’ (France). Remote Sens. Appl. Soc. Environ. 24, 100637 (2021)

    2021

  11. [19]

    & Debaine, F

    Lafitte, T., Robin, M., Launeau, P. & Debaine, F. Remote Sensing for Mapping Natura 2000 Habitats in the Brière Marshes: Setting Up a Long -Term Monitoring Strategy to Understand Changes. Remote Sens. 16, 2708 (2024)

    2024

  12. [20]

    Giannetti, F. et al. European Forest Types: toward an automated classification. Ann. For. Sci. 75, 6 (2018). 20

    2018

  13. [21]

    Immitzer, M. et al. Optimal Input Features for Tree Species Classification in Central Europe Based on Multi-Temporal Sentinel-2 Data. Remote Sens. 11, 2599 (2019)

    2019

  14. [22]

    Marzialetti, F. et al. Capturing Coastal Dune Natural Vegetation Types Using a Phenology-Based Mapping Approach: The Potential of Sentinel -2. Remote Sens. 11, 1506 (2019)

    2019

  15. [23]

    Huber, N. et al. Countrywide classification of permanent grassland habitats at high spatial resolution. Remote Sens. Ecol. Conserv. 9, 133–151 (2023)

    2023

  16. [24]

    Distribution and habitat suitability maps of revised EUNIS coastal and wetland habitats

    Hennekens, S. Distribution and habitat suitability maps of revised EUNIS coastal and wetland habitats

  17. [25]

    Distribution and habitat suitability maps of revised EU NIS Marine saltmarshes and Sparsely vegetated habitats

    Hennekens, S. Distribution and habitat suitability maps of revised EU NIS Marine saltmarshes and Sparsely vegetated habitats

  18. [26]

    Álvarez‐Martínez, J. M. et al. Modelling the area of occupancy of habitat types with remote sensing. Methods Ecol. Evol. 9, 580–593 (2018)

    2018

  19. [27]

    Jiménez‐Alfaro, B. et al. Modelling the distribution and compositional variation of plant communities at the continental scale. Divers. Distrib. 24, 978–990 (2018)

    2018

  20. [28]

    Chytrý, M. et al. EUNIS-ESy: Expert system for automatic classification of European vegetation plots to EUNIS habitats. Zenodo https://doi. org/10.5281/ZENODO.3841728 (2021)

    2021 doi

  21. [29]

    & Ozinga, W

    Schaminée, J., Hennekens, S. & Ozinga, W. The Dutch National Vegetation Database. Biodivers. Ecol. 4, 201–209 (2012)

    2012

  22. [30]

    NDFF Dutch vegetation database

    Hennekens, S. NDFF Dutch vegetation database. Dutch National Database of Flora and Fauna (NDFF) https://doi.org/10.15468/KSQXEP (2018)

    2018 doi

  23. [31]

    Données brutes, Campagnes annuelles 2005 et suivantes

    IGN – Inventaire Forestier National Français. Données brutes, Campagnes annuelles 2005 et suivantes. (2022)

    2022

  24. [32]

    MAES/EUNIS habitat map Austria 10m

    Umweltbundesamt. MAES/EUNIS habitat map Austria 10m. 56.1 MBytes PANGAEA https://doi.org/10.1594/PANGAEA.934147 (2021)

    2021 doi

  25. [33]

    & Grenouillet, G

    Buisson, L., Thuiller, W., Casajus, N., Lek, S. & Grenouillet, G. Uncertainty in ensemble forecasting of species distribution. Glob. Change Biol. 16, 1145–1157 (2010)

    2010

  26. [34]

    Biodiversity - Climate Change and the Ecologist

    Thuiller, W. Biodiversity - Climate Change and the Ecologist. Nature 448, 550–552 (2007)

    2007

  27. [35]

    Thuiller, W., Gueguen, M., Renaud, J., Karger, D. N. & Zimmermann, N. E. Uncertainty in ensembles of global biodiversity scenarios. Nat. Commun. 10, (2019)

    2019

  28. [36]

    & Zimmermann, N

    Guisan, A., Thuiller, W. & Zimmermann, N. E. Habitat Suitability and Distribution Models: With Applications in R . (Cambridge University Press, 2017). doi:10.1017/9781139028271

    2017 doi

  29. [37]

    & Hinton, G

    LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436–444 (2015)

    2015

  30. [38]

    Cheng, H.-T. et al. Wide & Deep Learning for Recommender Systems. in Proceedings of the 1st Workshop on Deep Learning for Recommender Systems 7–10 (ACM, Boston MA USA, 2016). doi:10.1145/2988450.2988454

    2016

  31. [39]

    Roberts, D. R. et al. Cross-validation st rategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography 40, 913–929 (2017)

    2017

  32. [40]

    Random forests

    Breiman, L. Random forests. Mach. Learn. 45, 5–32 (2001)

    2001

  33. [41]

    & Guestrin, C

    Chen, T. & Guestrin, C. XGBoost: A Scalable Tree Boosting System. in (2016)

    2016

  34. [42]

    Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V. & Gulin, A. CatBoost: unbiased boosting with categorical features. in Advances in Neural Information Processing Systems (eds. Bengio, S. et al.) vol. 31 (Curran Associates, Inc., 2018)

    2018

  35. [43]

    Ke, G. et al. LightGBM: A Highly Efficient Gradient Boosting Decision Tree. in Advances in Neural Information Processing Systems (eds. Guyon, I. et al.) vol. 30 (Curran Associates, Inc., 2017)

    2017

  36. [44]

    & Hinton, G

    LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436–444 (2015). 21

    2015

  37. [45]

    & Courville, A

    Goodfellow, I., Bengio, Y. & Courville, A. Deep Learning. (MIT Press, 2016)

    2016

  38. [46]

    T. -Y. Lin, P. Goyal, R. Girshick, K. He, & P. Dollár. Focal Loss for Dense Object Detection. in 2017 IEEE International Conference on Computer Vision (ICCV) 2999–3007 (2017). doi:10.1109/ICCV.2017.324

    2017 doi

  39. [47]

    Cao, Y., Larsen, D. P. & Thorne, R. S. -J. Rare species in multivariate analysis for bioassessment: some considerations. J. North Am. Benthol. Soc. 20, 144–153 (2001)

    2001

  40. [48]

    & Kajdanowicz, T

    Szymanski, P. & Kajdanowicz, T. Scikit-multilearn: a scikit-based Python environment for performing multi-label classification. J. Mach. Learn. Res. 20, 209–230

  41. [49]

    & Koyama, M

    Akiba, T., Sano, S., Yanase, T., Ohta, T. & Koyama, M. Optuna: A Next -generation Hyperparameter Optimization Framework. in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining 2623–2631 (Association for Computing Machinery, New York...

    2019

  42. [50]

    Olson, D. M. et al. Terrestrial ecoregions of th e world: a new map of life on Earth. Bioscience 51, 933–938 (2001). Tables Table 1. Number of vegetation plots for each EUNIS level 1 habitat formation Saltmarshes (MA2) Coastal (N) Wetlands (Q) Grassland (R) Shrubland (S) Fores...

    2001

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.