What is log loss and why is it important in classification?

Updated May 17, 2026

Short answer

Log loss measures the error between predicted probabilities and actual labels.

Deep explanation

Log loss, also called cross-entropy loss, penalizes incorrect confident predictions heavily. It evaluates how close predicted probabilities are to true labels. Lower log loss indicates better calibrated and more accurate probabilistic predictions. It is widely used in logistic regression and neural networks.

Real-world example

Used in Kaggle competitions for evaluating classification models.

Common mistakes

  • Confusing log loss with accuracy, ignoring probability quality.

Follow-up questions

  • Why does log loss penalize confident wrong predictions more?
  • How is log loss related to maximum likelihood?

More Supervised Learning interview questions

View all →