Isolation Forest is an unsupervised, tree-based anomaly detection methodology. See how each KernelSHAP and TreeSHAP can be utilized to elucidate its output.
Isolation Forest has turn out to be a staple in anomaly detection techniques [1]. Its benefit is having the ability to discover complicated anomalies in giant datasets with many options. Nonetheless, in relation to explaining these anomalies, this benefit rapidly turns into a weak point.
To take motion on an anomaly we regularly have to know the explanations for it being categorized as one. This perception is especially invaluable in real-world functions, resembling fraud detection, the place understanding the explanation behind an anomaly is commonly as essential as detecting it.
Sadly, with Isolation Forest, these explanations are hidden throughout the complicated mannequin construction. To uncover them, we flip to SHAP.
We are going to apply SHAP to IsolationForest and interpret its output. We are going to see that though that is an unsupervised mannequin we are able to nonetheless use SHAP to elucidate its anomaly scores. That’s to know:
- How options have contributed to the scores of particular person situations
- and which options are essential generally.
