machine learning framework: New Product Creates Computational
Models from Data
September 30, 2002--machine learning framework, now available from
Wolfram Research, Inc., is a flexible Mathematica application
package for business
and engineering professionals who want to extract computational models
from data.
With optimized, fuzzy logic-based machine learning methods and algorithms,
machine learning framework can be a powerful tool for all types of
data mining and machine learning applications:
- Media experts can extract features from images and signals to
detect defects.
- Process engineers can model chemical, metallurgical, or other
continuous processes to improve output.
- Process automation system developers can control complex
manufacturing machines.
- Marketing professionals can create profiles to improve the
positioning of their products and services.
- Financial engineers can analyze market characteristics to
calibrate models for financial instruments.
All methods in machine learning framework are implemented in C++
and are seamlessly integrated into Mathematica's intuitive computation,
visualization, and programming environment. Methods are highly
parameterized to ensure maximum flexibility and performance in terms of
computational speed and accuracy. Results can be easily visualized using
the Mathematica front end and can also be modified using
Mathematica programming to fine-tune the models.
Knowledge engineers and machine learning experts who want to customize,
configure, and integrate their own machine learning arrangements will
benefit from machine learning framework's open architecture and
Mathematica's high-level programming environment.
machine learning framework's online user's manual, which aids
with function description and command reference, is fully integrated into
the Mathematica Help Browser for convenience and ease of use. Also
available are online notebooks with descriptions of all methods and
templates for special predefined machine learning tasks.
machine learning framework is published and supported by Software
Competence Center Hagenberg (SCCH) GmbH and uni software plus GmbH.
Additional information is
available.
| 
