Course contentsShow
← Security
Lesson 2348 of 3,05245. Security Operations and MonitoringFree lesson

Baseline Establishment and Anomaly Detection

Creating behavioral baselines for systems and users to improve anomaly-based alert accuracy and reduce false positives.

Baseline Establishment and Anomaly Detection

You'll learn how creating behavioral baselines for systems and users dramatically improves anomaly-based alert accuracy and cuts down false positives.

The Core Problem

Your SIEM fires 500 alerts daily: "Unusual login time!" "High data transfer!" "New application launched!" But which are real threats? Without context, every deviation looks suspicious. That's where baselining transforms raw alerts into actionable intelligence.

What Is a Baseline?

A baseline is a detailed profile of normal behavior over time. Think of it like learning your car's typical fuel consumption—once you know you average 30 MPG, a sudden drop to 15 MPG signals something's wrong. Similarly, baselines capture:

  • User patterns: Login times, typical data volumes, geographic locations, applications accessed
  • System behaviors: CPU/memory usage, network traffic patterns, service dependencies, API call frequencies

Building Effective Baselines

  1. Collection Period: Gather data for 30–90 days minimum (avoiding holiday/incident anomalies)
  2. Granularity: Track metrics by time-of-day, day-of-week, user role, department
  3. Dynamic Updates: Baselines must evolve—summer vacation patterns differ from tax season workloads
  4. Multi-dimensional: Combine multiple metrics (time + volume + location) for stronger signal

Reducing False Positives

With baselines established, your SIEM rule evolves from "login at 2 AM = alert" to "login at 2 AM AND from unknown country AND accessing 10× normal data = high-priority alert." Engineers working late stops triggering noise; actual account compromise stands out.

The Anomaly Detection Advantage

Machine learning models can learn these baselines automatically, detecting subtle deviations humans might miss—like gradually escalating privilege usage or slowly increasing data exfiltration (boiling frog attacks).

Key Takeaway: Behavioral baselines transform security monitoring from reacting to every deviation into intelligently detecting meaningful anomalies, dramatically improving analyst efficiency and threat detection accuracy.