TreeNet® is Jerome Friedman's latest contribution to data mining and it promises to inaugurate a new era of exceptionally accurate predictive models. TreeNet® is a breakthrough technology that can offer surprisingly high accuracy while remaining remarkably robust in the presence of common data problems such as missing values and misrecorded training data. The technology is currently available for a variety of regression and classification problems.
TreeNet® wins major data mining competion.
In the FAQ document (click here to download the PDF), we provide further detail and the technology is covered in our advanced data mining seminars (see our training pages for dates and locations).
Technical Articles by Jerome Friedman are also available for download:
- Greedy Function Approximation: A Gradient Boosting Machine introduces the methodology.
- Stochastic Gradient Boosting discusses several improvements to the original idea.