文档介绍：-based decision tree model integrated with
GIS for assessment of soil heavy metals pollution
Wei Cheng
Department of Tourism, Resources and Environment
Zaozhuang University
Zaozhuang, China
Abstract—-based decision tree model was integrated with cover, and climate [7]. So integration of a decision tree
Geographical Information Systems (GIS) to assess the soil Cu approach with a GIS offers a potential solution in meeting this
pollution in Fuyang, Zhejiang, China. The integration of the challenge [6].
decision tree model with ArcGIS Engine using
implementation in Microsoft® Visual Basic provided an Coupling between models and GIS is described as loose,
approach for assessing the spatial distribution of soil Cu content tight, and embedded [8]. However, in most cases the GIS is
with high predictive accuracy. The decision tree model (CART) used to process and display inputs and outputs to an external
accuracy of assigning samples to the right Cu classes is % model that can be started automatically via scripting, or
and %, the Kappa coefficient is and manually (. loose coupling). This allows flexibility in the
respectively for training data and test data. This is a great modeling language, and does not require a model to be
paring with ordinary Kriging method for the rewritten, only simply modified to process