Spatial Data Modeller (SDM) is a collection of geoprocessing tools for adding categorical maps with interval, ordinal, or ratio scale maps to produce a predictive map of where something of interest is likely to occur. The tools include the data-driven methods of Weights of Evidence, Logistic Regression, and two supervised and one unsupervised neural network methods, and categorical tools for a knowledge-driven method Fuzzy Logic. These categorical fuzzification tools complement the early SDM Fuzzy Logic tools, now fully implemented in ArcGIS 10 (i.e., Spatial Overlay). All of the tools have help files that include references to publications discussing the applications of the methods implemented in the tool. Several of the tools create output rasters, tables, or files that require the user to enter a name.
This new release of SDM requires an active Spatial Analyst extension with ArcGIS 10.0.
Future SDM developments will continue to be implemented here at the University of Campinas by Prof. Carlos Roberto de Souza Filho and his research group. This site has been maintained by him since 2005 to contain historic versions of SDM and related documentation.
Highlights of the new SDM version:
The new Spatial Data Modeller (SDM) consists of two toolboxes for use with ArcGIS 10. The Demonstration Models Toolbox has models that demonstrate how to use the tools in the Spatial Data Modeller Toolbox. The tools in SDM include tools for Weights of Evidence, two neural nets, some tools to assign fuzzy memberships to categorical data for the Fuzzy Logic tools in the ArcGIS 10. and a set of useful utilities.
The user should read the enclosed Readme.pdf to learn how to install the files and use the tools in the toolbox. The first step is to extract the contents of the zip file to the system root directory, typically C:\. Then read the Readme.pdf file.
Also included is a set of data to use with the Demonstration Model Toolbox. This is most easily run by loading the MXD in the Carlin folder of the SDM folder. It is useful to successfully run and study the demonstration models as a foundation to running your own models.