GRNN2

GRNN_2: General Regression Neural Network - Anisotropic module

GRNN are a class of neural networks widely used for the continuous function mapping. They are based on a well known nonparametric (kernel) statistical estimators. An important advantage of the GRNN is that training is very fast.

You can download the software for free. The only payment we ask you is to send a postcard from your place to:

Vadim Timonin
Institute of Geomatics and Analysis of Risk (IGAR)
Amphipole, University of Lausanne
1015 Lausanne, Switzerland
Contact: mail to Vadim.Timonin@unil.ch

Copyright: IGAR, University of Lausanne, 2006

User manual:

• GRNN2 (anisotropic version)

• GRNN1 (isotropic version)

Publications:

• Lazy Mapping of environmental and pollution data

          (M. Kanevski)

• Spatial Predictions of Soil Contamination Using General Regression Neural Networks

          (M. Kanevski)

Download:

Used file formats:

If you have BDE (Borland Database Engine) installed, dBase files (*.dbf) can be used.
If no, use Geo-EAS data file format.

Standard Geo-EAS Data file.

It's ASCII (text) format. The first line of the file contains a descriptive title. The second line is the number of variables (NVAR) in the data file. Next NVAR lines (from line 3 to line NVAR+2) contain names for each variable (one line per variable). The data itself follows the header lines in NVAR columns (one for each variable) separated by space, tabulator or comma. Only numeric variables (integer or floating point) are supported.

The first few lines of a sample Geo-EAS format file are presented below: