Generalized Integrated Gradients: A practical method for explaining diverse ensembles
Abstract
We introduce Generalized Integrated Gradients (GIG), a formal extension of the Integrated Gradients (IG) (Sundararajan et al., 2017) method for attributing credit to the input variables of a predictive model. GIG improves IG by explaining a broader variety of functions that arise from practical applications of ML in domains like financial services. GIG is constructed to overcome limitations of Shapley (1953) and Aumann-Shapley (1974), and has desirable properties when compared to other approaches. We prove GIG is the only correct method, under a small set of reasonable axioms, for providing explanations for mixed-type models or games. We describe the implementation, and present results of experiments on several datasets and systems of models.
Cite
@article{arxiv.1909.01869,
title = {Generalized Integrated Gradients: A practical method for explaining diverse ensembles},
author = {John Merrill and Geoff Ward and Sean Kamkar and Jay Budzik and Douglas Merrill},
journal= {arXiv preprint arXiv:1909.01869},
year = {2019}
}
Comments
38 pages, submitted to JMLR 9/3/2019