False Negative

In binary classification, a false negative happens when the model fails to detect something that is actually present. For example, a security model may classify a malicious request as safe, or a medical model may fail to identify an existing condition. False negatives are commonly measured through metrics such as recall and sensitivity.

False negatives can be particularly costly when failing to identify a positive case creates significant consequences. Reducing them is critical in applications where missed threats, fraud, defects, or other important events can lead to harm or financial loss.