Multi-Task Learning with Summary Statistics

TL;DR

This paper introduces a flexible multi-task learning framework that uses summary statistics to overcome data-sharing constraints, with an adaptive parameter selection method and theoretical analysis applicable to real-world scenarios like healthcare.

Contribution

It proposes a novel multi-task learning approach utilizing summary statistics and an adaptive tuning method, supported by non-asymptotic theoretical analysis.

Findings

Effective performance demonstrated through extensive simulations.

Theoretical guarantees under various sample regimes.

Applicable to fields like genetic risk prediction.

Abstract

Multi-task learning has emerged as a powerful machine learning paradigm for integrating data from multiple sources, leveraging similarities between tasks to improve overall model performance. However, the application of multi-task learning to real-world settings is hindered by data-sharing constraints, especially in healthcare settings. To address this challenge, we propose a flexible multi-task learning framework utilizing summary statistics from various sources. Additionally, we present an adaptive parameter selection approach based on a variant of Lepski's method, allowing for data-driven tuning parameter selection when only summary statistics are available. Our systematic non-asymptotic analysis characterizes the performance of the proposed methods under various regimes of the sample complexity and overlap. We demonstrate our theoretical findings and the performance of the method through extensive simulations. This work offers a more flexible tool for training related models across various domains, with practical implications in genetic risk prediction and many other fields.