Search

Products

Smartware & Data Mining offers the following data-mining modules, which may be integrated together:


Binary classification model (cMODEL): supplies a statistical model for classifying data in two separate classes. For example, by creating a model of an industrial plant we can determine whether production will be greater (or less) than a given standard in real time.


 

Multi-class classification model (mcMODEL): an extension of the previous model to multi-category cases. The same example of the industrial plant applies here, but we may also distinguish between different quality levels..

Regression model (rMODEL): creates a statistical model for associating incoming data with numerical values. For instance, it can create a model reflecting traffic trends and forecast future trends.

Novelty detection or anomaly detection model (ocMODEL): a system which associates similar situations in large databases. For instance, a company can identify clusters of customers with similar needs in a customer database.

Rule extraction model (rMODEL): extraction of IF-THEN-ELSE  rules from statistical models produced using the modules described above in order to make relationships among data comprehensible and easy to understand for a human operator.

Dimensionality reduction model: applied in combination with one of the modules described above, it permits identification of significant variables that have made the greatest contributions to achieving results, reducing the number of dimensions of the problem.