Your Navigation Path
Case Base
MiningMart
Case
TemplateCase
Chain
IntegrateDifferentDataSources

Step: FeatureConstructionByAggregation

Name FeatureConstructionByAggregation
Belongs to Case TemplateCase
Uses Operator FeatureConstructionByRelation
Belongs to Chain IntegrateDifferentDataSources
Concepts of resulting Data Model CreditFraud
InhabitantsData
InputAssociationRules
PersonalData
WeatherData
ClientWithAggregatedInfo
Has Parameters AggregationOperator0
AggregationOperator1
TheOutputAttribute0
TheOutputAttribute1
TheOutputConcept
TheRelation
TheTargetAttribute0
TheTargetAttribute1

M4 Documentation:

This step is another example of how relationships can be used during preprocessing. The step extends the client concept by a new attribute whose values are computed from information in a related concept. Here, for each client, the frequency of buying the product that has been bought most often by all clients, is computed. Secondly, the frequency of buying a product from the product group that has been bought most often by all clients, is computed. Note that the information which product (group) has been bought most often in total is encoded in the relationship which is used as input to this operator.