F1 score is the weighted average of precision and recall. It considers both false positive and false negative values into account. It is used to measure a model’s performance.