Transfer Learning with Network Embeddings under Structured Missingness

TL;DR

This paper introduces TransNEST, a transfer learning framework that leverages network embeddings and structured prior information to improve data integration across heterogeneous sites with missing features.

Contribution

TransNEST is a novel method that incorporates hierarchical group structures and site-specific features into network embeddings, enhancing transfer learning under structured missingness.

Findings

TransNEST achieves better embedding accuracy than benchmarks.

It improves pediatric health data analysis from multi-site EHRs.

The method demonstrates strong finite-sample performance.

Abstract

Modern data-driven applications increasingly rely on large, heterogeneous datasets collected across multiple sites. Differences in data availability, feature representation, and underlying populations often induce structured missingness, complicating efforts to transfer information from data-rich settings to those with limited data. Many transfer learning methods overlook this structure, limiting their ability to capture meaningful relationships across sites. We propose TransNEST (Transfer learning with Network Embeddings under STructured missingness), a framework that integrates graphical data from source and target sites with prior group structure to construct and refine network embeddings. TransNEST accommodates site-specific features, captures within-group heterogeneity and between-site differences adaptively, and improves embedding estimation under partial feature overlap. We establish the convergence rate for the TransNEST estimator and demonstrate strong finite-sample performance in simulations. We apply TransNEST to a multi-site electronic health record study, transferring feature embeddings from a general hospital system to a pediatric hospital system. Using a hierarchical ontology structure, TransNEST improves pediatric embeddings and supports more accurate pediatric knowledge extraction, achieving the best accuracy for identifying pediatric-specific relational feature pairs compared with benchmark methods.