File	Date	Author	Commit
doc	2011-06-22	G. Kohring	[0f5b62] added a handler for Trial data
src	2011-06-22	G. Kohring	[0f5b62] added a handler for Trial data
LICENSE.txt	2011-06-08	G. Kohring	[8d0581] Initial commit prior to posting to SourceForge.
README	2011-06-08	G. Kohring	[8d0581] Initial commit prior to posting to SourceForge.
waf	2011-06-08	G. Kohring	[8d0581] Initial commit prior to posting to SourceForge.
wscript	2011-06-22	G. Kohring	[e40ca1] Merge branch 'develop'

Read Me

Stochastico

Version: 0.1.0

Stochastico is an open source implementation of stochastic discrimination (SD).
SD is a machine learning algorithm used for categorization, pattern recognition
and predictive modeling. It is the most general algorithm in the
ensemble learning class of algorithms.

This software is an early release alpha version.  See the src/test/resources
directory for a list of the data sets on which this software has been
successfully tested.

===========
Building
===========

To build you need to have Python installed on your machine.  Most modern Linux
versions have Python pre-installed.

To compile, type

./waf distclean
./waf configure
./waf build

This will configure then build the software. In the build directory you will
find two programs: stochastico and unit-tests.

Examples of parameter sets for various test cases can be found in the
directory src/test/resources.  Note: you will need to download the data
first and edit the properties file to point to the directory containing
the data.

To start, issue a command like:

./build/stochastico -p src/test/resource/bezdekIris.properties

Typical results are contained in the src/test/resource/results file.

Please provide feedback!  Positive or negative, whatever you have to say will
be useful!

Enjoy,
The Stochastico Team