Apply Association Rules (Detailed) (Operator Toolbox)
SynopsisThis operator allows you to get the detailed results of the application of association rules.
This operator allows you to check if a certain rule is true for given input sets. When running this operator, it will be checked for every rule in the rule set if they apply to a given example. If yes a new example in the output ExampleSet is generated. The information for this match includes the id, premise, conclusion as well as all available measures to assess the strength of the rule.
In contrast to Apply Association Rules, this operator can generate several matches for one input item, if more than one rule applies. In this case one can use proper post-processing to select the rule which matches most or return all matches.
To safe memory, this operator is only returning the id of an example, not all describing attributes. A new attribute of type integer is generated, if your example set does not contain an attribute of role id.
- input (Data Table)
Data set to be scored. This needs to be in the 'dummy coded format' as defined in FP-Growth.
- input (Association Rules)
The association rules generated by Generate Association Rule operator, which you want to apply.
- exa (Data Table)
An ExampleSet where each example represents one matching rule.
- ori (Data Table)
The original data set.
In this tutorial process we first build and apply association rules. Afterwards we post-process the resulting data in a way, that it is very similar to the normal apply association rules project.