Local Outlier Factor

LOF is a density-based unsupervised algorithm that detects anomalies by comparing a point's local density to that of its k-nearest neighbors.

The algorithm identifies outliers by calculating a Local Reachability Density (LRD) for each data point. It compares this value against the densities of its neighbors: a score significantly higher than 1.0 indicates the point is in a sparser region than its peers, marking it as a local outlier. This method outperforms global techniques (like standard deviation) in complex datasets where clusters have varying densities. It is a standard tool for detecting credit card fraud or network intrusions where anomalous behavior mimics legitimate patterns but lacks the same local consistency.

https://scikit-learn.org/stable/modules/outlier_detection.html#local-outlier-factor