You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently always splits annotations in train & test. This is best for judging classifier performance, but means that not all training input by the user is actually used to build the model.
Two alternatives:
Fuzzy mode. No test data, all training
Cross-validation: Slower (multi-run), but all training data is used for some models
The text was updated successfully, but these errors were encountered:
Currently always splits annotations in train & test. This is best for judging classifier performance, but means that not all training input by the user is actually used to build the model.
Two alternatives:
Fuzzy mode. No test data, all training
Cross-validation: Slower (multi-run), but all training data is used for some models
The text was updated successfully, but these errors were encountered: