Axiomatic Interpretability for Multiclass Additive Models

TL;DR

This paper extends GAMs to multiclass problems, introduces axioms for interpretability, and proposes a method to ensure models remain visually truthful without losing accuracy.

Contribution

It generalizes a GAM learning algorithm to multiclass, identifies interpretability axioms, and develops API to enforce these axioms on any multiclass additive model.

Findings

Multiclass GAMs outperform existing algorithms and match complex models in some cases.

API effectively enforces interpretability axioms on various multiclass additive models.

The approach maintains accuracy while improving interpretability in multiclass settings.

Abstract

Generalized additive models (GAMs) are favored in many regression and binary classification problems because they are able to fit complex, nonlinear functions while still remaining interpretable. In the first part of this paper, we generalize a state-of-the-art GAM learning algorithm based on boosted trees to the multiclass setting, and show that this multiclass algorithm outperforms existing GAM learning algorithms and sometimes matches the performance of full complexity models such as gradient boosted trees. In the second part, we turn our attention to the interpretability of GAMs in the multiclass setting. Surprisingly, the natural interpretability of GAMs breaks down when there are more than two classes. Naive interpretation of multiclass GAMs can lead to false conclusions. Inspired by binary GAMs, we identify two axioms that any additive model must satisfy in order to not be visually misleading. We then develop a technique called Additive Post-Processing for Interpretability (API), that provably transforms a pre-trained additive model to satisfy the interpretability axioms without sacrificing accuracy. The technique works not just on models trained with our learning algorithm, but on any multiclass additive model, including multiclass linear and logistic regression. We demonstrate the effectiveness of API on a 12-class infant mortality dataset.