WaKA: Data Attribution using K-Nearest Neighbors and Membership Privacy Principles

TL;DR

WaKA introduces a novel data attribution method based on k-nearest neighbors and Wasserstein principles, enabling efficient privacy risk assessment and data valuation without extensive sampling, bridging data attribution and membership inference.

Contribution

WaKA provides a unified, efficient framework for data attribution and membership inference using k-NN classifiers, improving computational efficiency and robustness over existing methods.

Findings

WaKA performs comparably to LiRA in membership inference tasks.

WaKA is more computationally efficient than LiRA.

WaKA is more robust than Shapley Values in data minimization tasks.

Abstract

In this paper, we introduce WaKA (Wasserstein K-nearest-neighbors Attribution), a novel attribution method that leverages principles from the LiRA (Likelihood Ratio Attack) framework and k-nearest neighbors classifiers (k-NN). WaKA efficiently measures the contribution of individual data points to the model's loss distribution, analyzing every possible k-NN that can be constructed using the training set, without requiring to sample subsets of the training set. WaKA is versatile and can be used a posteriori as a membership inference attack (MIA) to assess privacy risks or a priori for privacy influence measurement and data valuation. Thus, WaKA can be seen as bridging the gap between data attribution and membership inference attack (MIA) by providing a unified framework to distinguish between a data point's value and its privacy risk. For instance, we have shown that self-attribution values are more strongly correlated with the attack success rate than the contribution of a point to the model generalization. WaKA's different usage were also evaluated across diverse real-world datasets, demonstrating performance very close to LiRA when used as an MIA on k-NN classifiers, but with greater computational efficiency. Additionally, WaKA shows greater robustness than Shapley Values for data minimization tasks (removal or addition) on imbalanced datasets.