Classification basics

Model performance for regression

Binary Classification

Metrics (only for bin classification)

• Accuracy: the ratio of correct predictions to total predictions

• Precision: the ratio of correct predictions to the total number that it should have guessed

• Recall: The ratio of cases that are positive that are actually positive

• F1 score: Over all metic to describe precision and recall between (0 and 1, higher is better)

Classification metrics for more than two classes

Metrics:

• Confusion matrix: used to show what areas the model does not understand