Anomaly Detection Methods - Interactive Visualization

Anomaly detection finds data points that deviate significantly from expected patterns without requiring labeled examples. Statistical methods like Z-score flag points many standard deviations from the mean - fast but assumes normality. Isolation Forest randomly partitions the feature space and measures how quickly a point is isolated - anomalies are isolated in very few splits. Local Outlier Factor compares each point's density to its neighbors'.

• See Z-score flagging: points more than 3 standard deviations from the mean colored red - works perfectly for Gaussian data
• Watch isolation forest build random partition trees - anomalies have shorter average path lengths to isolation
• Understand LOF: a point is an outlier if its local density is much lower than its neighbors - finds local anomalies Z-score misses
• See how each method handles different anomaly types: point anomalies, contextual anomalies, and collective anomalies
• Learn when to use each method: Z-score for univariate monitoring, Isolation Forest for high-dimensional production data

Part of the EngineersOfAI Interactive 3D - free interactive visualizations covering every major concept in machine learning and AI engineering. Hover any element for a plain-English explanation. No code required.