Watch this video on YouTube
In the context of training an ML model to classify customer feedback into positive, negative, and neutral categories, which of the following steps would be most appropriate for evaluating model performance?