# Enhancing Decision Tree based Interpretation of Deep Neural Networks through L1-Orthogonal Regularization

## TL;DR

This paper introduces an L1-orthogonal regularization method that improves the interpretability of deep neural networks by enabling accurate and simple decision tree surrogates without sacrificing model performance.

## Contribution

The paper proposes a novel L1-orthogonal regularization technique that enhances the fidelity and simplicity of decision tree explanations for neural networks.

## Key findings

- L1-orthogonal regularization leads to lower complexity decision trees.
- The method maintains neural network accuracy while improving interpretability.
- Models with this regularization outperform other regularizers in fidelity.

## Abstract

One obstacle that so far prevents the introduction of machine learning models primarily in critical areas is the lack of explainability. In this work, a practicable approach of gaining explainability of deep artificial neural networks (NN) using an interpretable surrogate model based on decision trees is presented. Simply fitting a decision tree to a trained NN usually leads to unsatisfactory results in terms of accuracy and fidelity. Using L1-orthogonal regularization during training, however, preserves the accuracy of the NN, while it can be closely approximated by small decision trees. Tests with different data sets confirm that L1-orthogonal regularization yields models of lower complexity and at the same time higher fidelity compared to other regularizers.