Anomaly Detection

Anomaly DetectionDetectionThe capability to identify suspicious activity, attacks, or policy violations in time.: Detects unusual patterns that may indicate attacks or misconfigurations. This capability supports the early detection of suspicious activity. Effective results require high-quality data sourcesData SourceA system, sensor, log, or repository that supplies data for security analysis and decisions., tuned detection logicDetection LogicRules, queries, models, and conditions used to identify suspicious behavior., triage, and continuous optimizationContinuous OptimizationOngoing tuning of rules, processes, and resources using measured operational results..

How it works and where it fits

Anomaly Detection connects data sources to detection or assessment logic. Raw records become security-relevant only when timing, identity, system context, and expected behavior are considered. Rules, correlations, statistical models, and analyst decisions may work together; no single method reliably covers every attack pattern.

Practical security relevance

Operational quality is reflected in coverage, data completeness, detection time, and false-alert workload. Data sources need owners, time synchronization, retention, and quality controls. Detections should be tested, versioned, and improved using real incidents. Every meaningful alert also requires triage guidance, escalation, and possible response actions.

Related concepts

• Behavioral AnalyticsBehavioral AnalyticsAnalyzes user and system behavior to detect suspicious deviations.: Analyzes user and system behavior to detect suspicious deviations.
• User and Entity Behavior AnalyticsUser and Entity Behavior AnalyticsDetects anomalous behavior by users, devices, and services using analytics.: Detects anomalous behavior by users, devices, and services using analytics.
• Detection EngineeringDetection EngineeringSystematic development, testing, and maintenance of rules for attack detection.: Systematic development, testing, and maintenance of rules for attack detection.
• Threat HuntingThreat HuntingSearches for previously undetected attacker activity based on hypotheses.: Searches for previously undetected attacker activity based on hypotheses.