MiningApplyTask (MAT) is an object which defines how the model is applied to data. Its execution results in the creation of a data set in the database. This data contain a specified set of columns, each of which contains the selected statistics for each observation, appropriate to the type of the applied model (probability of class membership, segment number, etc.).
Table 12.10. Connected objects
| Status | Name | Description | Type | Reference |
|---|---|---|---|---|
| included | directMapping | an object which determines which columns from the scored set should appear in the result set. The selected columns will be copied to the result set | Elements | own |
| included | modelAssignment | an object used for mapping the variables onto model signature | AttributeAssignmentSetImpl | own |
| obligatory | applyOutput | an object which stores the specification of the scoring results; depends on the type of the model (classification output type, time series output type, etc.) | MiningApplyOutput | own |
| obligatory | sourceData | the data for scoring | PhysicalData | by name |
| obligatory | targetData | the set which will store the scoring results | PhysicalData | by name |
| obligatory | model | a model in the Metadata Repository, created as the outcome of MBT | MiningModel | by name |