Does it make sense to use the train, test, and validation model using the Association Rules Technique?
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Finding association rules is an unsupervised learning task (or exploratory task). You don't actually know which rules you want to find before you actually find them, so there's nothing to test against. Validating against a separated instance set is usually done on supervised learning tasks such as classification and regression.
