Label-shift robust federated feature screening for high-dimensional classification

TL;DR

This paper proposes a novel federated feature screening method, LR-FFS, that is robust to label shift, computationally efficient, and effective in high-dimensional, heterogeneous data environments.

Contribution

It introduces a unified framework for federated feature screening under label shift and develops LR-FFS, a method that remains robust without increasing computational costs.

Findings

LR-FFS outperforms existing methods in diverse client environments.

The method effectively controls false discovery rate in federated settings.

Experimental results confirm robustness against label shift and model misspecification.

Abstract

Distributed and federated learning are important tools for high-dimensional classification of large datasets. To reduce computational costs and overcome the curse of dimensionality, feature screening plays a pivotal role in eliminating irrelevant features during data preprocessing. However, data heterogeneity, particularly label shifting across different clients, presents significant challenges for feature screening. This paper introduces a general framework that unifies existing screening methods and proposes a novel utility, label-shift robust federated feature screening (LR-FFS), along with its federated estimation procedure. The framework facilitates a uniform analysis of methods and systematically characterizes their behaviors under label shift conditions. Building upon this framework, LR-FFS leverages conditional distribution functions and expectations to address label shift without adding computational burdens and remains robust against model misspecification and outliers. Additionally, the federated procedure ensures computational efficiency and privacy protection while maintaining screening effectiveness comparable to centralized processing. We also provide a false discovery rate (FDR) control method for federated feature screening. Experimental results and theoretical analyses demonstrate LR-FFS's superior performance across diverse client environments, including those with varying class distributions, sample sizes, and missing categorical data.