Isolation Forest

Isolation Forest detects anomalies by randomly partitioning data using decision trees. Anomalous data points are easier to isolate because they are few and different from normal observations.

It scores each point based on how quickly it gets separated from the rest of the data.

• Isolation Forest is computationally efficient and works well with high-dimensional data. It does not require labeled data, making it ideal for discovering unusual patterns you did not know to look for.
• Identify and remove outlier data points before training other ML models for cleaner results.

Use Cases

• Outlier geometry detection: Find building elements with abnormal dimensions (e.g., unusually small rooms or oversized walls).
• Quality control: Detect geometric anomalies in a batch of elements that may indicate modeling errors.