Multiclass classification

Hi everyone. the task is predicting 4 classes from dataset which contains 1000 feature and 11000 rows. after finding the suitable learner and hyperparameters we need to predict labels with another test dataset (with same number of features) which is not labeled.
If i use the PCA block the number of features may increase or decrease. but I need to predict classes for another data set which is not labeled and got the error that some features in test data set ( not labeled one) is not found.
I really appreciate if anyone can guide me

@Amirrezasafari welcome to the KNIME forum. You might take a look at this example about a multi-class problem and how to evaluate it. Make sure to also read the additional “External Ressources” - especially the threads marked [multiclass]

1 Like

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.