Severity classification is a method for assigning a measured condition to levels, such as low, moderate, or high, based on its frequency, intensity, or other defined indicators. In wearable health software, it may summarize patterns in sensor data-for example, breathing disruptions detected during sleep. A severity level is an estimate produced under specific rules, not necessarily a clinical diagnosis.
In cyber security, severity classification helps teams prioritize alerts and incidents by estimated impact or urgency, so higher-risk issues receive attention first. It can also be used in health systems to guide users toward further evaluation. Classification depends on measurement quality, thresholds, and validation; noisy data or poorly chosen cutoffs can lead to misleading results. Defenders and product teams should communicate what each level means, validate classifications against appropriate evidence, and avoid treating an automated category as proof of a condition or threat.