Why does the classification work differently on untrained documents after adding a new class?

Question

When a new document class is added into the classifier, for the invoices that had not been trained, why does the classifier wrongly classify these documents, though previously they were classified correctly?

Answer

It is expected behavior that new classes and new trained documents can affect classification results as the new model is generated. In this case, add the wrongly classified documents to the document set, assign correct classes, and re-train the skill.

Have more questions? Submit a request

Comments

0 comments

Please sign in to leave a comment.