Techniques to Improve Ecological Interpretability of Black-Box Machine Learning Models View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2021-10-28

AUTHORS

Thomas Welchowski, Kelly O. Maloney, Richard Mitchell, Matthias Schmid

ABSTRACT

Statistical modeling of ecological data is often faced with a large number of variables as well as possible nonlinear relationships and higher-order interaction effects. Gradient boosted trees (GBT) have been successful in addressing these issues and have shown a good predictive performance in modeling nonlinear relationships, in particular in classification settings with a categorical response variable. They also tend to be robust against outliers. However, their black-box nature makes it difficult to interpret these models. We introduce several recently developed statistical tools to the environmental research community in order to advance interpretation of these black-box models. To analyze the properties of the tools, we applied gradient boosted trees to investigate biological health of streams within the contiguous USA, as measured by a benthic macroinvertebrate biotic index. Based on these data and a simulation study, we demonstrate the advantages and limitations of partial dependence plots (PDP), individual conditional expectation (ICE) curves and accumulated local effects (ALE) in their ability to identify covariate–response relationships. Additionally, interaction effects were quantified according to interaction strength (IAS) and Friedman’s H2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H^2$$\end{document} statistic. Interpretable machine learning techniques are useful tools to open the black-box of gradient boosted trees in the environmental sciences. This finding is supported by our case study on the effect of impervious surface on the benthic condition, which agrees with previous results in the literature. Overall, the most important variables were ecoregion, bed stability, watershed area, riparian vegetation and catchment slope. These variables were also present in most identified interaction effects. In conclusion, graphical tools (PDP, ICE, ALE) enable visualization and easier interpretation of GBT but should be supported by analytical statistical measures. Future methodological research is needed to investigate the properties of interaction tests. Supplementary materials accompanying this paper appear on-line. More... »

PAGES

175-197

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/s13253-021-00479-7

DOI

http://dx.doi.org/10.1007/s13253-021-00479-7

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1142240872


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/01", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Mathematical Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0104", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Statistics", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany", 
          "id": "http://www.grid.ac/institutes/grid.10388.32", 
          "name": [
            "Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Welchowski", 
        "givenName": "Thomas", 
        "id": "sg:person.01072763633.64", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01072763633.64"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "U.S. Geological Survey (USGS) Eastern Ecological Science Center at the Leetown Research Laboratory, 11649, Leetown Rd, 25430, Kearneysville, WV, USA", 
          "id": "http://www.grid.ac/institutes/grid.2865.9", 
          "name": [
            "U.S. Geological Survey (USGS) Eastern Ecological Science Center at the Leetown Research Laboratory, 11649, Leetown Rd, 25430, Kearneysville, WV, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Maloney", 
        "givenName": "Kelly O.", 
        "id": "sg:person.014030242625.51", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014030242625.51"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "U.S. Environmental Protection Agency Office of Water Washington, 20460, Washington, DC, USA", 
          "id": "http://www.grid.ac/institutes/grid.418698.a", 
          "name": [
            "U.S. Environmental Protection Agency Office of Water Washington, 20460, Washington, DC, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Mitchell", 
        "givenName": "Richard", 
        "id": "sg:person.015320662343.00", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015320662343.00"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany", 
          "id": "http://www.grid.ac/institutes/grid.10388.32", 
          "name": [
            "Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Schmid", 
        "givenName": "Matthias", 
        "id": "sg:person.0745220772.74", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0745220772.74"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/s10750-008-9655-y", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1009225584", 
          "https://doi.org/10.1007/s10750-008-9655-y"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1023/a:1003442013650", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1038558926", 
          "https://doi.org/10.1023/a:1003442013650"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s10651-010-0158-4", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1046581893", 
          "https://doi.org/10.1007/s10651-010-0158-4"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-0-387-84858-7", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1032573094", 
          "https://doi.org/10.1007/978-0-387-84858-7"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1023/a:1010933404324", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1024739340", 
          "https://doi.org/10.1023/a:1010933404324"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-540-87479-9_34", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1005796247", 
          "https://doi.org/10.1007/978-3-540-87479-9_34"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1186/1471-2105-8-25", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1019863657", 
          "https://doi.org/10.1186/1471-2105-8-25"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s11634-016-0276-4", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1046420813", 
          "https://doi.org/10.1007/s11634-016-0276-4"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2021-10-28", 
    "datePublishedReg": "2021-10-28", 
    "description": "Statistical modeling of ecological data is often faced with a large number of variables as well as possible nonlinear relationships and higher-order interaction effects. Gradient boosted trees (GBT) have been successful in addressing these issues and have shown a good predictive performance in modeling nonlinear relationships, in particular in classification settings with a categorical response variable.  They also tend to be robust against outliers. However, their black-box nature makes it difficult to interpret these models. We introduce several recently developed statistical tools to the environmental research community in order to advance interpretation of these black-box models. To analyze the properties of the tools, we applied gradient boosted trees to investigate biological health of streams within the contiguous USA, as measured by a benthic macroinvertebrate biotic index. Based on these data and a simulation study, we demonstrate the advantages and limitations of partial dependence plots (PDP), individual conditional expectation (ICE) curves and accumulated local effects (ALE) in their ability to identify covariate\u2013response relationships. Additionally, interaction effects were quantified according to interaction strength (IAS) and Friedman\u2019s H2\\documentclass[12pt]{minimal}\n\t\t\t\t\\usepackage{amsmath}\n\t\t\t\t\\usepackage{wasysym}\n\t\t\t\t\\usepackage{amsfonts}\n\t\t\t\t\\usepackage{amssymb}\n\t\t\t\t\\usepackage{amsbsy}\n\t\t\t\t\\usepackage{mathrsfs}\n\t\t\t\t\\usepackage{upgreek}\n\t\t\t\t\\setlength{\\oddsidemargin}{-69pt}\n\t\t\t\t\\begin{document}$$H^2$$\\end{document} statistic. Interpretable machine learning techniques are useful tools to open the black-box of gradient boosted trees in the environmental sciences. This finding is supported by our case study on the effect of impervious surface on the benthic condition, which agrees with previous results in the literature. Overall, the most important variables were ecoregion, bed stability, watershed area, riparian vegetation and catchment slope. These variables were also present in most identified interaction effects. In conclusion, graphical tools (PDP, ICE, ALE) enable visualization and easier interpretation of GBT but should be supported by analytical statistical measures. Future methodological research is needed to investigate the properties of interaction tests.\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Supplementary materials accompanying this paper appear on-line.", 
    "genre": "article", 
    "id": "sg:pub.10.1007/s13253-021-00479-7", 
    "inLanguage": "en", 
    "isAccessibleForFree": true, 
    "isPartOf": [
      {
        "id": "sg:journal.1134206", 
        "issn": [
          "1085-7117", 
          "1537-2693"
        ], 
        "name": "Journal of Agricultural, Biological and Environmental Statistics", 
        "publisher": "Springer Nature", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "1", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "27"
      }
    ], 
    "keywords": [
      "categorical response variables", 
      "partial dependence plots", 
      "covariate-response relationships", 
      "interaction strength", 
      "Black-box Machine Learning Models", 
      "black-box models", 
      "ecological interpretability", 
      "statistical modeling", 
      "higher order interaction effects", 
      "expectation curve", 
      "statistical tools", 
      "nonlinear relationship", 
      "statistical measures", 
      "simulation study", 
      "response variables", 
      "black-box nature", 
      "environmental research community", 
      "machine learning models", 
      "supplementary material", 
      "good predictive performance", 
      "previous results", 
      "easy interpretation", 
      "future methodological research", 
      "graphical tool", 
      "classification setting", 
      "predictive performance", 
      "interpretable machine", 
      "possible nonlinear relationship", 
      "methodological research", 
      "model", 
      "variables", 
      "large number", 
      "statistics", 
      "gradient", 
      "outliers", 
      "properties", 
      "environmental science", 
      "learning model", 
      "modeling", 
      "ecological data", 
      "interaction effects", 
      "interpretability", 
      "tool", 
      "technique", 
      "GBT", 
      "macroinvertebrate biotic index", 
      "useful tool", 
      "interpretation", 
      "stability", 
      "case study", 
      "order", 
      "curves", 
      "number", 
      "research community", 
      "important variables", 
      "machine", 
      "trees", 
      "performance", 
      "advantages", 
      "data", 
      "conditions", 
      "science", 
      "Friedman", 
      "results", 
      "effect", 
      "local effects", 
      "nature", 
      "surface", 
      "lines", 
      "limitations", 
      "slope", 
      "strength", 
      "literature", 
      "bed stability", 
      "streams", 
      "materials", 
      "plots", 
      "measures", 
      "contiguous USA", 
      "issues", 
      "relationship", 
      "visualization", 
      "setting", 
      "study", 
      "index", 
      "area", 
      "research", 
      "test", 
      "ability", 
      "watershed area", 
      "conclusion", 
      "interaction test", 
      "impervious surfaces", 
      "catchment slope", 
      "benthic conditions", 
      "community", 
      "USA", 
      "findings", 
      "riparian vegetation", 
      "vegetation", 
      "biotic indices", 
      "biological health", 
      "ecoregions", 
      "health", 
      "paper"
    ], 
    "name": "Techniques to Improve Ecological Interpretability of Black-Box Machine Learning Models", 
    "pagination": "175-197", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1142240872"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/s13253-021-00479-7"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1007/s13253-021-00479-7", 
      "https://app.dimensions.ai/details/publication/pub.1142240872"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2022-05-20T07:39", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220519/entities/gbq_results/article/article_903.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "https://doi.org/10.1007/s13253-021-00479-7"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/s13253-021-00479-7'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/s13253-021-00479-7'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s13253-021-00479-7'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s13253-021-00479-7'


 

This table displays all metadata directly associated to this object as RDF triples.

222 TRIPLES      22 PREDICATES      138 URIs      122 LITERALS      6 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/s13253-021-00479-7 schema:about anzsrc-for:01
2 anzsrc-for:0104
3 schema:author N578d33e3233e406b991969ab952d0535
4 schema:citation sg:pub.10.1007/978-0-387-84858-7
5 sg:pub.10.1007/978-3-540-87479-9_34
6 sg:pub.10.1007/s10651-010-0158-4
7 sg:pub.10.1007/s10750-008-9655-y
8 sg:pub.10.1007/s11634-016-0276-4
9 sg:pub.10.1023/a:1003442013650
10 sg:pub.10.1023/a:1010933404324
11 sg:pub.10.1186/1471-2105-8-25
12 schema:datePublished 2021-10-28
13 schema:datePublishedReg 2021-10-28
14 schema:description Statistical modeling of ecological data is often faced with a large number of variables as well as possible nonlinear relationships and higher-order interaction effects. Gradient boosted trees (GBT) have been successful in addressing these issues and have shown a good predictive performance in modeling nonlinear relationships, in particular in classification settings with a categorical response variable. They also tend to be robust against outliers. However, their black-box nature makes it difficult to interpret these models. We introduce several recently developed statistical tools to the environmental research community in order to advance interpretation of these black-box models. To analyze the properties of the tools, we applied gradient boosted trees to investigate biological health of streams within the contiguous USA, as measured by a benthic macroinvertebrate biotic index. Based on these data and a simulation study, we demonstrate the advantages and limitations of partial dependence plots (PDP), individual conditional expectation (ICE) curves and accumulated local effects (ALE) in their ability to identify covariate–response relationships. Additionally, interaction effects were quantified according to interaction strength (IAS) and Friedman’s H2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H^2$$\end{document} statistic. Interpretable machine learning techniques are useful tools to open the black-box of gradient boosted trees in the environmental sciences. This finding is supported by our case study on the effect of impervious surface on the benthic condition, which agrees with previous results in the literature. Overall, the most important variables were ecoregion, bed stability, watershed area, riparian vegetation and catchment slope. These variables were also present in most identified interaction effects. In conclusion, graphical tools (PDP, ICE, ALE) enable visualization and easier interpretation of GBT but should be supported by analytical statistical measures. Future methodological research is needed to investigate the properties of interaction tests.      Supplementary materials accompanying this paper appear on-line.
15 schema:genre article
16 schema:inLanguage en
17 schema:isAccessibleForFree true
18 schema:isPartOf N008fa0bf58104a03b248fc7612fb5cc7
19 Na506265b689040f6aaa2872e9ae0344c
20 sg:journal.1134206
21 schema:keywords Black-box Machine Learning Models
22 Friedman
23 GBT
24 USA
25 ability
26 advantages
27 area
28 bed stability
29 benthic conditions
30 biological health
31 biotic indices
32 black-box models
33 black-box nature
34 case study
35 catchment slope
36 categorical response variables
37 classification setting
38 community
39 conclusion
40 conditions
41 contiguous USA
42 covariate-response relationships
43 curves
44 data
45 easy interpretation
46 ecological data
47 ecological interpretability
48 ecoregions
49 effect
50 environmental research community
51 environmental science
52 expectation curve
53 findings
54 future methodological research
55 good predictive performance
56 gradient
57 graphical tool
58 health
59 higher order interaction effects
60 impervious surfaces
61 important variables
62 index
63 interaction effects
64 interaction strength
65 interaction test
66 interpretability
67 interpretable machine
68 interpretation
69 issues
70 large number
71 learning model
72 limitations
73 lines
74 literature
75 local effects
76 machine
77 machine learning models
78 macroinvertebrate biotic index
79 materials
80 measures
81 methodological research
82 model
83 modeling
84 nature
85 nonlinear relationship
86 number
87 order
88 outliers
89 paper
90 partial dependence plots
91 performance
92 plots
93 possible nonlinear relationship
94 predictive performance
95 previous results
96 properties
97 relationship
98 research
99 research community
100 response variables
101 results
102 riparian vegetation
103 science
104 setting
105 simulation study
106 slope
107 stability
108 statistical measures
109 statistical modeling
110 statistical tools
111 statistics
112 streams
113 strength
114 study
115 supplementary material
116 surface
117 technique
118 test
119 tool
120 trees
121 useful tool
122 variables
123 vegetation
124 visualization
125 watershed area
126 schema:name Techniques to Improve Ecological Interpretability of Black-Box Machine Learning Models
127 schema:pagination 175-197
128 schema:productId N06877ac220794316a6572769d8998b55
129 Nc13811354f244423a39f17731d7966d0
130 schema:sameAs https://app.dimensions.ai/details/publication/pub.1142240872
131 https://doi.org/10.1007/s13253-021-00479-7
132 schema:sdDatePublished 2022-05-20T07:39
133 schema:sdLicense https://scigraph.springernature.com/explorer/license/
134 schema:sdPublisher N8d4916acd6b94cea9782b295bfa8e04b
135 schema:url https://doi.org/10.1007/s13253-021-00479-7
136 sgo:license sg:explorer/license/
137 sgo:sdDataset articles
138 rdf:type schema:ScholarlyArticle
139 N008fa0bf58104a03b248fc7612fb5cc7 schema:issueNumber 1
140 rdf:type schema:PublicationIssue
141 N06877ac220794316a6572769d8998b55 schema:name doi
142 schema:value 10.1007/s13253-021-00479-7
143 rdf:type schema:PropertyValue
144 N578d33e3233e406b991969ab952d0535 rdf:first sg:person.01072763633.64
145 rdf:rest Nd265b25d49224ea3a3fb78ccba1bb2f0
146 N8d4916acd6b94cea9782b295bfa8e04b schema:name Springer Nature - SN SciGraph project
147 rdf:type schema:Organization
148 Na506265b689040f6aaa2872e9ae0344c schema:volumeNumber 27
149 rdf:type schema:PublicationVolume
150 Nc13811354f244423a39f17731d7966d0 schema:name dimensions_id
151 schema:value pub.1142240872
152 rdf:type schema:PropertyValue
153 Nd265b25d49224ea3a3fb78ccba1bb2f0 rdf:first sg:person.014030242625.51
154 rdf:rest Ne05b1009e4724a318bebfb892eb41941
155 Nd414923f0a224f0f83ab1c6e14d3c478 rdf:first sg:person.0745220772.74
156 rdf:rest rdf:nil
157 Ne05b1009e4724a318bebfb892eb41941 rdf:first sg:person.015320662343.00
158 rdf:rest Nd414923f0a224f0f83ab1c6e14d3c478
159 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
160 schema:name Mathematical Sciences
161 rdf:type schema:DefinedTerm
162 anzsrc-for:0104 schema:inDefinedTermSet anzsrc-for:
163 schema:name Statistics
164 rdf:type schema:DefinedTerm
165 sg:journal.1134206 schema:issn 1085-7117
166 1537-2693
167 schema:name Journal of Agricultural, Biological and Environmental Statistics
168 schema:publisher Springer Nature
169 rdf:type schema:Periodical
170 sg:person.01072763633.64 schema:affiliation grid-institutes:grid.10388.32
171 schema:familyName Welchowski
172 schema:givenName Thomas
173 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01072763633.64
174 rdf:type schema:Person
175 sg:person.014030242625.51 schema:affiliation grid-institutes:grid.2865.9
176 schema:familyName Maloney
177 schema:givenName Kelly O.
178 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014030242625.51
179 rdf:type schema:Person
180 sg:person.015320662343.00 schema:affiliation grid-institutes:grid.418698.a
181 schema:familyName Mitchell
182 schema:givenName Richard
183 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.015320662343.00
184 rdf:type schema:Person
185 sg:person.0745220772.74 schema:affiliation grid-institutes:grid.10388.32
186 schema:familyName Schmid
187 schema:givenName Matthias
188 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0745220772.74
189 rdf:type schema:Person
190 sg:pub.10.1007/978-0-387-84858-7 schema:sameAs https://app.dimensions.ai/details/publication/pub.1032573094
191 https://doi.org/10.1007/978-0-387-84858-7
192 rdf:type schema:CreativeWork
193 sg:pub.10.1007/978-3-540-87479-9_34 schema:sameAs https://app.dimensions.ai/details/publication/pub.1005796247
194 https://doi.org/10.1007/978-3-540-87479-9_34
195 rdf:type schema:CreativeWork
196 sg:pub.10.1007/s10651-010-0158-4 schema:sameAs https://app.dimensions.ai/details/publication/pub.1046581893
197 https://doi.org/10.1007/s10651-010-0158-4
198 rdf:type schema:CreativeWork
199 sg:pub.10.1007/s10750-008-9655-y schema:sameAs https://app.dimensions.ai/details/publication/pub.1009225584
200 https://doi.org/10.1007/s10750-008-9655-y
201 rdf:type schema:CreativeWork
202 sg:pub.10.1007/s11634-016-0276-4 schema:sameAs https://app.dimensions.ai/details/publication/pub.1046420813
203 https://doi.org/10.1007/s11634-016-0276-4
204 rdf:type schema:CreativeWork
205 sg:pub.10.1023/a:1003442013650 schema:sameAs https://app.dimensions.ai/details/publication/pub.1038558926
206 https://doi.org/10.1023/a:1003442013650
207 rdf:type schema:CreativeWork
208 sg:pub.10.1023/a:1010933404324 schema:sameAs https://app.dimensions.ai/details/publication/pub.1024739340
209 https://doi.org/10.1023/a:1010933404324
210 rdf:type schema:CreativeWork
211 sg:pub.10.1186/1471-2105-8-25 schema:sameAs https://app.dimensions.ai/details/publication/pub.1019863657
212 https://doi.org/10.1186/1471-2105-8-25
213 rdf:type schema:CreativeWork
214 grid-institutes:grid.10388.32 schema:alternateName Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany
215 schema:name Department of Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Venusberg-Campus 1, 53127, Bonn, Germany
216 rdf:type schema:Organization
217 grid-institutes:grid.2865.9 schema:alternateName U.S. Geological Survey (USGS) Eastern Ecological Science Center at the Leetown Research Laboratory, 11649, Leetown Rd, 25430, Kearneysville, WV, USA
218 schema:name U.S. Geological Survey (USGS) Eastern Ecological Science Center at the Leetown Research Laboratory, 11649, Leetown Rd, 25430, Kearneysville, WV, USA
219 rdf:type schema:Organization
220 grid-institutes:grid.418698.a schema:alternateName U.S. Environmental Protection Agency Office of Water Washington, 20460, Washington, DC, USA
221 schema:name U.S. Environmental Protection Agency Office of Water Washington, 20460, Washington, DC, USA
222 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...