Skip to Main Content
U.S. Forest Service
Caring for the land and serving people

United States Department of Agriculture

Home > Search > Publication Information

  1. Share via EmailShare on FacebookShare on LinkedInShare on Twitter
    Dislike this pubLike this pub
    Author(s): E. FreemanG. Moisen; J. Coulston; B. Wilson
    Date: 2014
    Source: The International Forestry Review. 16(5): 408.
    Publication Series: Abstract
    Station: Rocky Mountain Research Station
    PDF: Download Publication  (73.0 KB)


    Random forests (RF) and stochastic gradient boosting (SGB), both involving an ensemble of classification and regression trees, are compared for modeling tree canopy cover for the 2011 National Land Cover Database (NLCD). The objectives of this study were twofold. First, sensitivity of RF and SGB to choices in tuning parameters was explored. Second, performance of the two final models was compared by assessing the importance of, and interaction between, predictor variables, the global accuracy metrics derived from an independent test set, and the visual quality of the resultant maps of tree canopy cover. Examination of relative variable importance elucidated the differences in how RF and SGB make use of correlated predictor variables. SGB had a tendency to concentrate variable importance in fewer variables, whereas RF tended to spread importance out amongst more variables. The predictive accuracy of RF and SGB was remarkably similar on all four of the pilot regions, by all the accuracy measures examined. RF is simpler to implement than SGB, as RF both has fewer parameters needing tuning, and also was less sensitive to these parameters.

    Publication Notes

    • You may send email to to request a hard copy of this publication.
    • (Please specify exactly which publication you are requesting and your mailing address.)
    • We recommend that you also print this page and attach it to the printout of the article, to retain the full citation information.
    • This article was written and prepared by U.S. Government employees on official time, and is therefore in the public domain.


    Freeman, E.; Moisen, G.; Coulston, J.; Wilson, B. 2014. Random forests and stochastic gradient boosting for predicting tree canopy cover: Comparing tuning processes and model performance. The International Forestry Review. 16(5): 408.


    random forests (RF), stochastic gradient boosting (SGB), tree canopy cover

    Related Search

    XML: View XML
Show More
Show Fewer
Jump to Top of Page