A table used to evaluate a classification model's performance by comparing predicted labels against actual labels.
A confusion matrix breaks down exactly where a model is right and wrong — true positives, false positives, true negatives, false negatives — giving a much fuller picture than accuracy alone.
