TIBCO Patterns Machine Learning Platform

The TIBCO Patterns Machine Learning Platform maximizes the difference between a match (good set of scores) and a non-match (bad set of scores). It does this using a Learn model that has been created and trained for the particular application. A model can be created and trained using the Learn UI (see the TIBCO Patterns Learn UI Guide), the Learn API library (see the TIBCO Patterns Learn API reference documentation), or TIBCO Patterns can create a Learn model for you. The model can be used to evaluate input feature vectors. It can also be used to predict the match score in a search operation. To use the model you must first load the model into memory with a given name. The model is then specified by name as a parameter for the search operation or feature vector evaluation operation. The structure of the query or the input feature vector used with the model must be identical to those used to train the model.