A common scenario in creating Mining Model is to predict their accuracy by running them against a test data set which is different from the training data. This prevents the problem of over fitting and gives a better measure of the accuracy of the generated models.
Read More - Register for Free Membership
