Windows Neural Networks (WinNN) is a shareware windows neural networks simulator. WinNN can handle multi-layered feed forward networks and train using modified back-propagation. WinNN has very friendly user interface and fast computational engine to run the calculations. WinNN can import data, train, and plot the results.

WinNN runs under Windows 95/98/Me/NT/2000/XP

Some of the features of WinNN:

• WinNN32 is a fast 32-bit implementation that works under Windows 95/98/NT/2000/XP.
• Smoothly trains in the background. Supports DDE with Excel (use NN predictions in Excel).
• Built in graphics to plot the NN architecture and the results.
• All files are written in simple ASCII format that can be used by other programs.
• The trained NN can easily be used from all programming languages.
• This is a mature program that is being used for since 1994.

Links and references to work done using WinNN:

Soil degradation and crop productivity research for conservation technology development...
Use of Neural Networks to Determine Properties of Alkanes from their 13C-NMR Spectra
Predicting Force Constants Using Neural Networks
Determination of dilution of lignite during selective....
P. Salem (editor),"Neural Networks Applications in Social Science Research.", Organizational communication and change. Pp. 151-173. Cresskill, NJ: Hampton Press (1999).