Explicating feature contribution using Random Forest proximity distances

TL;DR

This paper introduces a method to quantify feature contributions in Random Forests by analyzing proximity distances, enabling better interpretability, auditing, and error analysis of black-box models.

Contribution

The paper presents a novel approach to explicate feature contributions in Random Forests using proximity distances, fully specifying the process for binary features.

Findings

Allows calculation of feature contributions in decision-making

Enables auditing and explanation of black-box decisions

Facilitates post-hoc error analysis

Abstract

In Random Forests, proximity distances are a metric representation of data into decision space. By observing how changes in input map to the movement of instances in this space we are able to determine the independent contribution of each feature to the decision-making process. For binary feature vectors, this process is fully specified. As these changes in input move particular instances nearer to the in-group or out-group, the independent contribution of each feature can be uncovered. Using this technique, we are able to calculate the contribution of each feature in determining how black-box decisions were made. This allows explication of the decision-making process, audit of the classifier, and post-hoc analysis of errors in classification.