Detection outcomes or classification outcomes, come from the field of binary classification.
There are applied in vulnerability management to describe how accurately a vulnerability scanning tool or process identifies issues.
For example, if a vulnerability scanner identifies a specific security flaw (such as an outdated software version with known exploits), and this flaw actually exists on the system, thatβs a true positive. It indicates the detection was accurate and actionable.
4 Outcomes#
Outcomes | Definition | Vuln Exists | Vuln Not_Found |
---|---|---|---|
True Positive | Correctly identifies a vulnerability that exists. | βοΈ | |
False Positive | Incorrectly identifies a vulnerability that doesn’t exist. | βοΈ | |
True Negative | Correctly identifies that a vulnerability doesn’t exist. | βοΈ | |
False Negative | Incorrectly fails to identify a vulnerability that exists. | βοΈ |
In machine learning and cybersecurity, these outcomes are sometimes also referred to as the confusion matrix outcomes, as they are part of the confusion matrix used to evaluate the performance of a classifier.