Skip to yearly menu bar Skip to main content


Poster
in
Workshop: 3rd Workshop on Interpretable Machine Learning in Healthcare (IMLH)

Efficient Estimation of Local Robustness of Machine Learning Models

Tessa Han · Suraj Srinivas · Himabindu Lakkaraju

Keywords: [ Classification ] [ debugging ] [ bias ] [ robustness ]


Abstract:

Machine learning models often need to be robust to noisy input data. The effect of real-world noise (which is often random) on model predictions is captured by a model’s local robustness, i.e., the consistency of model predictions in a local region around an input. Local robustness is therefore an important characterization of real-world model behavior and can be useful for debugging models and establishing user trust. However, the naïve approach to computing local robustness based on Monte-Carlo sampling is statistically inefficient, leading to prohibitive computational costs for large-scale applications. In this work, we develop the first analytical estimators to efficiently compute local robustness of multi-class discriminative models using local linear function approximation and the multivariate Normal CDF. Through the derivation of these estimators, we show how local robustness is connected to concepts such as randomized smoothing and softmax probability. We also confirm empirically that these estimators accurately and efficiently compute the local robustness of standard deep learning models. In addition, we demonstrate these estimators’ usefulness for various tasks involving local robustness, such as measuring robustness bias and identifying examples that are vulnerable to noise perturbation in a dataset. y developing analytical estimators of local robustness, but also makes its computation practical, enabling the use of local robustness in critical downstream applications.

Chat is not available.