Download
Hengl-et-al_2018_Random forest as a generic framework for predictive modeling of spatial and spatio-temporal variables.pdf 7,35MB
WeightNameValue
1000 Titel
  • Random forest as a generic framework for predictive modeling of spatial and spatio-temporal variables
1000 Autor/in
  1. Hengl, Tomislav |
  2. Nussbaum, Madlene |
  3. Wright, Marvin N. |
  4. Heuvelink, Gerard B.M. |
  5. Gräler, Benedikt |
1000 Erscheinungsjahr 2018
1000 LeibnizOpen
1000 Art der Datei
1000 Publikationstyp
  1. Artikel |
1000 Online veröffentlicht
  • 2018-08-29
1000 Erschienen in
1000 Quellenangabe
  • 6: e5518
1000 FRL-Sammlung
1000 Copyrightjahr
  • 2018
1000 Lizenz
1000 Verlagsversion
  • http://dx.doi.org/10.7717/peerj.5518 |
1000 Ergänzendes Material
  • https://peerj.com/articles/5518/#supplemental-information |
1000 Publikationsstatus
1000 Begutachtungsstatus
1000 Sprache der Publikation
1000 Abstract/Summary
  • Random forest and similar Machine Learning techniques are already used to generate spatial predictions, but spatial location of points (geography) is often ignored in the modeling process. Spatial auto-correlation, especially if still existent in the cross-validation residuals, indicates that the predictions are maybe biased, and this is suboptimal. This paper presents a random forest for spatial predictions framework (RFsp) where buffer distances from observation points are used as explanatory variables, thus incorporating geographical proximity effects into the prediction process. The RFsp framework is illustrated with examples that use textbook datasets and apply spatial and spatio-temporal prediction to numeric, binary, categorical, multivariate and spatiotemporal variables. Performance of the RFsp framework is compared with the state-of-the-art kriging techniques using fivefold cross-validation with refitting. The results show that RFsp can obtain equally accurate and unbiased predictions as different versions of kriging. Advantages of using RFsp over kriging are that it needs no rigid statistical assumptions about the distribution and stationarity of the target variable, it is more flexible towards incorporating, combining and extending covariates of different types, and it possibly yields more informative maps characterizing the prediction error. RFsp appears to be especially attractive for building multivariate spatial prediction models that can be used as “knowledge engines” in various geoscience fields. Some disadvantages of RFsp are the exponentially growing computational intensity with increase of calibration data and covariates and the high sensitivity of predictions to input data quality. The key to the success of the RFsp framework might be the training data quality—especially quality of spatial sampling (to minimize extrapolation problems and any type of bias in data), and quality of model validation (to ensure that accuracy is not effected by overfitting). For many data sets, especially those with lower number of points and covariates and close-to-linear relationships, model-based geostatistics can still lead to more accurate predictions than RFsp.
1000 Sacherschließung
lokal R statistical computing
lokal Spatiotemporal data
lokal Spatial data
lokal Geostatistics
lokal Predictive modeling
lokal Random forest
lokal Pedometrics
lokal Sampling
lokal Kriging
1000 Fachgruppe
  1. Medizin |
  2. Gesundheitswesen |
1000 Fächerklassifikation (DDC)
1000 Liste der Beteiligten
  1. https://frl.publisso.de/adhoc/creator/SGVuZ2wsIFRvbWlzbGF2|https://frl.publisso.de/adhoc/creator/TnVzc2JhdW0sIE1hZGxlbmU=|https://frl.publisso.de/adhoc/creator/V3JpZ2h0LCBNYXJ2aW4gTi4=|https://frl.publisso.de/adhoc/creator/SGV1dmVsaW5rLCBHZXJhcmQgQi5NLg==|https://frl.publisso.de/adhoc/creator/R3LDpGxlciwgQmVuZWRpa3Q=
1000 Förderer
  1. German Federal Ministry for Economic Affairs and Energy |
1000 Fördernummer
  1. 50EE1715C
1000 Förderprogramm
  1. -
1000 Dateien
  1. Hengl-et-al_2018_Random forest as a generic framework for predictive modeling of spatial and spatio-temporal variables
1000 Förderung
  1. 1000 joinedFunding-child
    1000 Förderer German Federal Ministry for Economic Affairs and Energy |
    1000 Förderprogramm -
    1000 Fördernummer 50EE1715C
1000 Objektart article
1000 Beschrieben durch
1000 @id frl:6411685.rdf
1000 Erstellt am 2018-12-10T12:41:47.921+0100
1000 Erstellt von 266
1000 beschreibt frl:6411685
1000 Bearbeitet von 122
1000 Zuletzt bearbeitet Wed Dec 12 12:50:46 CET 2018
1000 Objekt bearb. Wed Dec 12 12:50:29 CET 2018
1000 Vgl. frl:6411685
1000 Oai Id
  1. oai:frl.publisso.de:frl:6411685 |
1000 Sichtbarkeit Metadaten public
1000 Sichtbarkeit Daten public
1000 Gegenstand von

View source