Software

I am pleased to announce that the WinMine Toolkit version 1.0, developed
by the Machine Learning and Applied Statistics group in Microsoft
Research, is now available for academic use. This release includes tools
for

-	Learning Bayesian networks, dependency networks, and decision
	trees from data
-	Evaluating the predictive accuracy of learned models
-	Viewing and browsing learned models

This release runs under the Windows NT/2000 operating systems.  It
includes command-line executables that facilitate scripting of
experiments.  The toolkit uses XML formats for both data and models.

To learn more about the features of this release or to download it,
please go to

http://research.microsoft.com/~dmax/WinMine/tooldoc.htm

Max Chickering