Learning Qualitatively Diverse and Interpretable Rules for Classification
Abstract
There has been growing interest in developing accurate models that can also be explained to humans. Unfortunately, if there exist multiple distinct but accurate models for some dataset, current machine learning methods are unlikely to find them: standard techniques will likely recover a complex model that combines them. In this work, we introduce a way to identify a maximal set of distinct but accurate models for a dataset. We demonstrate empirically that, in situations where the data supports multiple accurate classifiers, we tend to recover simpler, more interpretable classifiers rather than more complex ones.
Cite
@article{arxiv.1806.08716,
title = {Learning Qualitatively Diverse and Interpretable Rules for Classification},
author = {Andrew Slavin Ross and Weiwei Pan and Finale Doshi-Velez},
journal= {arXiv preprint arXiv:1806.08716},
year = {2018}
}
Comments
Presented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden (revision fixes minor issues)