Prediction-Oriented Transfer Learning for Survival Analysis

2026-03-11

AI Summary

This work addresses the limitations of conventional transfer learning in survival analysis when target event data are scarce, which typically relies on the strong assumption of parameter similarity between source and target under a Cox model and requires access to individual-level source data. To overcome these constraints, we propose a prediction-oriented transfer learning framework that bypasses modeling the source data distribution altogether and instead transfers predictive knowledge to enhance survival prediction on the target domain—without requiring access to or sharing of source data. Built upon a flexible semiparametric transformation model, our approach integrates an EM algorithm with a specially designed prediction-transfer regularization term to achieve stable and efficient estimation. Theoretical analysis establishes a faster convergence rate for the proposed estimator, and both simulation studies and real-world analysis of breast cancer data demonstrate its superior predictive accuracy over existing methods.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

Transfer learning is beneficial for survival analysis, especially when the target study has a limited number of events. However, existing transfer learning methods rely on the restrictive assumption that the target and source studies share similar parameters under Cox models, and most require access to individual-level source data. In this article, we propose a novel transfer learning framework that enhances model-based survival prediction by transferring predictive rather than distributional knowledge from source studies. Our approach employs flexible semiparametric transformation models for the target data while eliminating the need to model or share the source data. The ingeniously designed penalty enables simple and stable computation via an EM algorithm. We rigorously establish the asymptotic properties of the proposed estimator and show that it achieves a faster convergence rate than the target-only estimator when source knowledge is sufficiently accurate. We demonstrate the advantages of our methods through extensive simulation studies and an application to two major breast cancer studies.

Problem

• transfer learning
• survival analysis
• prediction
• Cox models
• source data

Innovation

• prediction-oriented transfer learning
• survival analysis
• semiparametric transformation models
• EM algorithm
• asymptotic convergence