Principle:Interpretml Interpret Tree Interpretation
Metadata
| Field | Value |
|---|---|
| Sources | Interpretml_Interpret |
| Domains | Interpretability, Feature_Attribution |
| Updated | 2026-02-07 |
Overview
Tree interpretation decomposes predictions of tree-based models into per-feature contributions by tracing decision paths from root to leaf.
Description
TreeInterpreter wraps the treeinterpreter package to provide local feature contribution explanations for scikit-learn tree-based models. For each prediction, it traces the decision path through the tree and attributes the change in prediction at each node to the feature used for splitting. This produces a decomposition where the prediction equals the bias (average training response) plus the sum of individual feature contributions. The approach is model-specific to decision trees and random forests.
Usage
Use tree interpretation when you need fast, exact local explanations for scikit-learn decision tree or random forest models. Requires the treeinterpreter package.