Software

Salford Systems Releases TreeNet(tm) 2.0

No Cost 30-day Download Available at
www.salford-systems.com/treenetreg.php

TreeNet has established a remarkable record of superior performance over the past four years, including winning several data mining competitions. In 2004, one of our clients used TreeNet to win the "Most Accurate" category in the KDD2004 competition and, using TreeNet, Salford won all four categories of the 2002 NCR Teradata/Duke University Churn Modeling competition. The news has been even better in industrial practice, where TreeNet is at the core of recent high performance credit card transactions fraud detection systems, credit scoring applications, targeted marketing, and text mining. In financial applications TreeNet has been shown to be definitively superior to other technologies across a broad range of predictive models. In text mining and e-commerce, the advantages of TreeNet have been shown to be even more dramatic.

TreeNet 2.0 builds on this demonstrated record by adding improved reporting and new graphics, and by simplifying the model construction process. The software now makes automatic model building, rapid feature selection, and accelerated score card development even easier. In addition built-in cross validation now makes TreeNet suitable for analyzing data with a small number of records or with a small number of records of interest. Finally, models can now be exported to Java, as well as to PMML, C, and other popular programnming and statistical package languages.

To arrange for an evaluation license, to see user stories, or to learn more about this remarkable new technology created by Stanford University Professor Jerome Friedman, please contact us at:

abaldwin@salford-systems.com