How could I apply a score dataset ( dataset with no values of a target variable) onto a model in KNIME for predictive models?

Dear All,

I was wondering, if anybody know how to apply a score dataset to a data mining model in KNIME.

I can easy use train and test datasets for classification accuracy of a data mining model, for example decision tree or a bayes classifiers, in KNIME. But when apply a score dataset (in testing box) with no values of target attribute, KNIME gives a zero classification.

I have a score dataset but do not know if there is a way to apply the score dataset to a DM model in KNIME software. Unfortunately, I can not skip the score dataset part and only relay on train and test accuracy of my model.

So, any help will be welcomed.



Hi Seydan, please have look to the first example here that shows how to learn, apply and score a dataset (divided into train and test data) using a Decision Tree model. Regards, Thomas

Hi Thomas,

Thanks for your email. I have looked at the examples.

But as I mentioned on my first email. I could divide a dataset in to training and testing using the partition icon and apply them to a decision tree model or any predictive model. Infact I sucessfully used training and testing datasets for prediction.

But when it comes to applying the score dataset to predict unknown data, I could know find a way to apply the score set which contains unknown target attribute values.

I provided decision tree predictor with the score set, I receive 0 classification accuracy. I think this is due to not having a target attribute values.

I am not expert on KNIME and was wondering if somebody could provide an information how to resolve the issue.

There is nothing wrong with the datasets or the idea of using score set to predict the unknown target values. (classification in data mining).



Hi Seyhan,

Did you get solution for the above problem? Could you please share with us.