Scalable and Distributed Individualized Treatment Rules for Massive Datasets

TL;DR

This paper introduces a scalable, privacy-preserving distributed method for learning individualized treatment rules using a convolution-smoothed SVM, achieving optimal performance with minimal communication across data sources.

Contribution

It develops a novel convolution-smoothed weighted SVM and an efficient distributed learning algorithm that preserves data privacy and reduces communication costs for multi-source ITRs.

Findings

Method achieves optimal statistical performance with fixed communication rounds.

Efficient coordinate gradient descent guarantees linear convergence.

Validated through simulations and real sepsis treatment data.

Abstract

Synthesizing information from multiple data sources is crucial for constructing accurate individualized treatment rules (ITRs). However, privacy concerns often present significant barriers to the integrative analysis of such multi-source data. Classical meta-learning, which averages local estimates to derive the final ITR, is frequently suboptimal due to biases in these local estimates. To address these challenges, we propose a convolution-smoothed weighted support vector machine for learning the optimal ITR. The accompanying loss function is both convex and smooth, which allows us to develop an efficient multi-round distributed learning procedure for ITRs. Such distributed learning ensures optimal statistical performance with a fixed number of communication rounds, thereby minimizing coordination costs across data centers while preserving data privacy. Our method avoids pooling subject-level raw data and instead requires only sharing summary statistics. Additionally, we develop an efficient coordinate gradient descent algorithm, which guarantees at least linear convergence for the resulting optimization problem. Extensive simulations and an application to sepsis treatment across multiple intensive care units validate the effectiveness of the proposed method.