Towards More Efficient Data Valuation in Healthcare Federated Learning using Ensembling

TL;DR

This paper introduces SaFE, an efficient method for calculating Shapley values in healthcare federated learning, enabling accurate contribution assessment without prohibitive computational costs.

Contribution

The paper presents SaFE, a novel ensembling-based technique for precise Shapley value computation in healthcare federated learning, improving over existing approximation methods.

Findings

SaFE closely approximates exact Shapley values.

SaFE outperforms current approximation methods in accuracy.

Applicable in medical imaging with heterogeneous data.

Abstract

Federated Learning (FL) wherein multiple institutions collaboratively train a machine learning model without sharing data is becoming popular. Participating institutions might not contribute equally, some contribute more data, some better quality data or some more diverse data. To fairly rank the contribution of different institutions, Shapley value (SV) has emerged as the method of choice. Exact SV computation is impossibly expensive, especially when there are hundreds of contributors. Existing SV computation techniques use approximations. However, in healthcare where the number of contributing institutions are likely not of a colossal scale, computing exact SVs is still exorbitantly expensive, but not impossible. For such settings, we propose an efficient SV computation technique called SaFE (Shapley Value for Federated Learning using Ensembling). We empirically show that SaFE computes values that are close to exact SVs, and that it performs better than current SV approximations. This is particularly relevant in medical imaging setting where widespread heterogeneity across institutions is rampant and fast accurate data valuation is required to determine the contribution of each participant in multi-institutional collaborative learning.