Explaining Machine Learning Models using Entropic Variable Projection
Preprint
Abstract
In this paper, we present a new explainability formalism designed to shed light on how each input variable of a test set impacts the predictions of machine learning models. Hence, we propose a group explainability formalism for trained machine learning decision rules, based on their response to the variability of the input variables distribution. In order to emphasize the impact of each input variable, this formalism uses an information theory framework that quantifies the influence of all input-output observations based on entropic projections. This is thus the first unified and model agnostic formalism enabling data scientists to interpret the dependence between the input variables, their impact on the prediction errors, and their influence on the output predictions. Convergence rates of the entropic projections are provided in the large sample case. Most importantly, we prove that computing an explanation in our framework has a low algorithmic complexity, making it scalable to real-life large datasets. We illustrate our strategy by explaining complex decision rules learned by using XGBoost, Random Forest or Deep Neural Network classifiers on various datasets such as Adult Income, MNIST, CelebA, Boston Housing, Iris, as well as synthetic ones. We finally make clear its differences with the explainability strategies LIME and SHAP, that are based on single observations. Results can be reproduced by using the freely distributed Python toolbox https://gems-ai.aniti.fr/.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- The Computational Complexity of Understanding Binary Classifier DecisionsJournal of Artificial Intelligence Research, 2021
- Methods for interpreting and understanding deep neural networksDigital Signal Processing, 2018
- Geodesic PCA versus Log-PCA of Histograms in the Wasserstein SpaceSIAM Journal on Scientific Computing, 2018
- A Gaussian Process Regression Model for Distribution InputsIEEE Transactions on Information Theory, 2017
- On Shapley Value for Measuring Importance of Dependent InputsSIAM/ASA Journal on Uncertainty Quantification, 2017
- On the rate of convergence in Wasserstein distance of the empirical measureProbability Theory and Related Fields, 2014
- Density modification-based reliability sensitivity analysisJournal of Statistical Computation and Simulation, 2014
- A Survey on Transfer LearningIEEE Transactions on Knowledge and Data Engineering, 2009
- Theoretical Comparison between the Gini Index and Information Gain CriteriaAnnals of Mathematics and Artificial Intelligence, 2004
- The role of trust in automation relianceInternational Journal of Human-Computer Studies, 2003