Instruction: Explain what a confusion matrix is and how it is used.
Context: This question evaluates the candidate's knowledge of confusion matrices as a tool for evaluating the performance of classification models.
Official answer available
Preview the opening of the answer, then unlock the full walkthrough.
The way I'd explain it in an interview is this: A confusion matrix is a table that compares predicted classes against actual classes so you can see where the model is getting things right and, more importantly, what kinds of mistakes it is...