FedSCS-XGB -- Federated Server-centric surrogate XGBoost for continual health monitoring

TL;DR

This paper introduces FedSCS-XGB, a federated server-centric surrogate XGBoost protocol for continuous health monitoring using wearable sensors, achieving near-centralized performance in human activity recognition tasks.

Contribution

It presents a novel distributed XGBoost-based protocol inspired by PAX, with theoretical convergence guarantees and empirical validation on wearable sensor data.

Findings

Achieves less than 1% performance gap compared to centralized XGBoost.

The protocol converges to equivalent solutions under proper conditions.

Demonstrates effectiveness on real-world wearable sensor datasets.

Abstract

Wearable sensors with local data processing can detect health threats early, enhance documentation, and support personalized therapy. In the context of spinal cord injury (SCI), which involves risks such as pressure injuries and blood pressure instability, continuous monitoring can help mitigate these by enabling early deDtection and intervention. In this work, we present a novel distributed machine learning (DML) protocol for human activity recognition (HAR) from wearable sensor data based on gradient-boosted decision trees (XGBoost). The proposed architecture is inspired by Party-Adaptive XGBoost (PAX) while explicitly preserving key structural and optimization properties of standard XGBoost, including histogram-based split construction and tree-ensemble dynamics. First, we provide a theoretical analysis showing that, under appropriate data conditions and suitable hyperparameter selection, the proposed distributed protocol can converge to solutions equivalent to centralized XGBoost training. Second, the protocol is empirically evaluated on a representative wearable-sensor HAR dataset, reflecting the heterogeneity and data fragmentation typical of remote monitoring scenarios. Benchmarking against centralized XGBoost and IBM PAX demonstrates that the theoretical convergence properties are reflected in practice. The results indicate that the proposed approach can match centralized performance up to a gap under 1\% while retaining the structural advantages of XGBoost in distributed wearable-based HAR settings.