Interpretable and Differentially Private Predictions
This addresses the privacy risks for users when interpretable models reveal sensitive data characteristics, though it appears incremental as it builds on existing differential privacy and interpretability techniques.
The paper tackles the problem of balancing interpretability and privacy in machine learning models, proposing a family of simple models that approximate complex ones using locally linear maps per class to provide high classification accuracy and differentially private explanations. The approach is demonstrated on several image benchmark datasets and a medical dataset.
Interpretable predictions, where it is clear why a machine learning model has made a particular decision, can compromise privacy by revealing the characteristics of individual data points. This raises the central question addressed in this paper: Can models be interpretable without compromising privacy? For complex big data fit by correspondingly rich models, balancing privacy and explainability is particularly challenging, such that this question has remained largely unexplored. In this paper, we propose a family of simple models in the aim of approximating complex models using several locally linear maps per class to provide high classification accuracy, as well as differentially private explanations on the classification. We illustrate the usefulness of our approach on several image benchmark datasets as well as a medical dataset.