Phase where the model's performance is evaluated?

• A. Model assessment
• B. Model training
• C. Data collection and preprocessing
• D. Model selection

Answer: (A)

The phase that checks how well the model actually performs on new, unseen data is model assessment. After training, you test the model on a held-out dataset and compute metrics suitable for the task (such as accuracy, precision/recall, F1, or ROC-AUC). This evaluation reveals how well it generalizes beyond the training data and whether it meets the required performance before deployment. Data collection and preprocessing is about getting and cleaning the data, model training is the learning process, and model selection involves choosing among candidate models or configurations—with the performance check happening during the assessment step.