What is accuracy in model evaluation?

Updated May 17, 2026

Short answer

Accuracy measures the proportion of correct predictions over total predictions.

Deep explanation

Accuracy = (TP + TN) / (TP + TN + FP + FN). It works well for balanced datasets but can be misleading for imbalanced ones.

Real-world example

Spam detection systems often report accuracy but rely on precision/recall for deeper evaluation.

Common mistakes

  • Using accuracy for heavily imbalanced datasets.

Follow-up questions

  • When is accuracy misleading?
  • What metric should replace it?

More Model Evaluation interview questions

View all →