Confusion Matrix
Domain: Evaluation
🎯 Motivation
A table showing correct and incorrect predictions broken down by category. Reveals where the model confuses classes.
📋 Concrete Example
Matrix: Cat->Cat: 85 (correct), Cat->Dog: 10 (confused), Dog->Cat: 5, Dog->Dog: 90. The model sometimes confuses cats with dogs. Shows specific weaknesses.
🔗 Analogy
A teacher’s error analysis - ‘Student A: good at algebra, struggles with geometry. Student B: good at essays, struggles with grammar.’ The confusion matrix shows where the model gets confused.