2
$\begingroup$

If I have a large set of transactions where in each I buy a set of goods and I want to do market basket analysis using either A-priori or FP Growth or any other data mining method, you typically get an antecedent, consequent, support confidence and lift scores after running the algorithm. However, compared to traditional supervised machine learning algorithms like Linear Regression, Random Forests, Gradient Boost, etc there does not appear to be a corresponding methodology where you split into a train and test set, train on the train dataset and check for cross validation on the test set for algorithms such as A-priori and FP Growth. How do you know that your model is truly good, are there some other metrics that you know that you are not overfitting your model and that it has no bias?

$\endgroup$
1
  • $\begingroup$ any help please? $\endgroup$ Aug 15, 2017 at 15:24

1 Answer 1

1
$\begingroup$

I am facing the same situation. I think that the reason is that it is not a supervised model, so there is not a pre-established label. If you don't have a label you can't get precision, recall or RMSE so it does not make any sense to do cross-validation or to split data.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.