Hello Larry02744,
There are a variety of algorithms and source code to do this kind of
processing available. The following sites provide a selection of both.
Weka 3 - machine learning software in Java
http://www.cs.waikato.ac.nz/~ml/weka/
Released under the GNU Public License, it implements algorithms for
classification, numeric prediction, clustering, and meta schemes
(e.g., regression via classification). From reading descriptions of
other products, this may be the most comprehensive. This site also
includes links to a number of related sites for machine learning.
Data Mining using Decision Trees
http://planetsourcecode.com/vb/scripts/ShowCode.asp?txtCodeId=504&lngWId=5
A free visual basic application to extract information from a variety
of database systems. A couple related applications using pareto trees
and "one rule" data mining are at
http://www.msaccessguru.com/visual_basic_data_mining.htm
Graf-fx data mining
http://www.gr-fx.com/graf-fx.htm
Implements a number of algorithms using Microsoft Access. It is a
shareware utility and full source code is available for a fee.
Open Source Index of Data Mining Software
http://dmoz.org/Computers/Software/Databases/Data_Mining/Public_Domain_Software/
A comprehensive list of software products that are available, but does
not describe the algorithms.
Potter's Wheel for Data Transformation
http://control.cs.berkeley.edu/abc/
An interesting description of methods to enter transformations and
analyze data. Source code for Windows platforms.
A few projects with good ideas but not yet providing software.
Code Web: Data Mining of Software
http://codeweb.sourceforge.net/
An effort to get useful object oriented software from publically
available sources. Unfortunately, no source code is yet available for
this project. There is however a paper describing some of the
techniques. There is a link to another site at...
http://www.dwsimpson.com/dataminingsoftware.html
with a number of good references.
Continuous monitoring of remote data sources
http://www.cs.cornell.edu/database/himalaya/
The idea behind Himalaya is to mine data as it is generated. An
example used is data gathered from SNMP data collection (used to
manage a network). Currently implements a few algorithms but has plans
for many more. Source code section is empty.
Several more good references can be found with a search using phrases
such as
source code data mining
If there is a specific database, platform, or language you wish to
use, add that phrase to the search to improve the search results.
--Maniac |