Log Loss
Log Loss, also called logarithmic loss or cross-entropy loss, measures how well a classification model's predicted probabilities match the actual outcomes. It penalizes incorrect predictions, especially when the model is highly confident in the wrong answer.
What is Log Loss?
Log Loss evaluates the probability assigned to the correct class for each prediction. A model receives a lower loss when it assigns a high probability to the correct outcome and a higher loss when it assigns a low probability to it. Unlike accuracy, Log Loss considers the confidence of the model's predictions.
Why is Log Loss Important?
Log Loss helps assess the quality and calibration of probabilistic classification models. It provides more information than simple accuracy because it distinguishes between confident and uncertain predictions. Lower Log Loss generally indicates better probabilistic predictions.
Common use cases
Log Loss is commonly used to evaluate binary and multiclass classification models, particularly when predicted probabilities are important for decision-making, ranking, or risk estimation.