·        Support Vector Machines (SVM) Modeling has been recently developed from Machine Learning/Statistical disciplines and has proven to outperform other state-of-the-art statistical modeling techniques, such as logistic regression and neural networks.

·        Joe Weissmann, President, a recognized industry leader in DM modeling applications, has been involved in predictive techniques for the DM industry for 30+ years  –  including projects for a number of Fortune 500 firms and top DM practitioners,  and has introduced a number of innovations along the way.

·        Padhraic Smyth, Chief Scientist, is a world renowned authority in the field and has lectured and been published extensively, including with MIT Press. Padhraic has led scientific teams at NASA’s Jet Propulsion Laboratory and has taught some of the pioneers in the field, including PhD’s who have implemented cutting edge text categorization methods at Yahoo! and Google.

·        SVM algorithms now routinely categorize millions of text documents for companies such as Microsoft, IBM and major Web companies; and are utilized in face recognition, robotics , optical character recognition, and gene expression data classification, to name a few applications.

·        SVM Modeling avoids the pitfall of overfitting of data. This is a major reason Neural Networks in the 90s have been an industry disappointment.

·        One reason for SVM success is that SVMs focus on prediction of unseen data rather than perfecting weights on the modeled data as the other techniques do. This seems obvious enough but hasn't been delivered until now!