I want evaluate a model. I have train set and test set. I worked on train set and create the best model on this. After that, i scored this model with the test set.
Now, how i can evaluate this model? with which data set exactly? with train set or i have split the train set to create a validation set? Can you explain me in detail please?
You can evaluate a deployed model on a new dataset using the Evaluate recipe: https://academy.dataiku.com/latest/tutorial/machine-learning/model-lifecycle.html#monitoring-a-model...
Alternatively, you can specify your test set at the model training stage with the 'explicit extracts' mode: https://doc.dataiku.com/dss/latest/machine-learning/supervised/settings.html#explicit-extracts