Rank-based transfer learning for high-dimensional survival data with application to sepsis data

2025-04-15

AI Summary

To address the high mortality and scarcity of clinical survival data in MSSA sepsis, this paper proposes the first rank-based transfer learning framework tailored for transformer models. Methodologically, it employs C-index–driven source data selection, a two-stage transfer mechanism (transfer followed by debiasing), Lasso-regularized survival regression, bootstrap resampling, and asymptotic normality theory to construct coefficient confidence intervals. Theoretical contributions include a falsifiable criterion for source data identification, ℓ₁/ℓ₂ estimation error bounds, detection consistency, and proof of asymptotic validity for confidence intervals. Empirically, on MIMIC-IV sepsis data, the method achieves an average 8.2% improvement in C-index. Simulation studies confirm that estimation error converges at the theoretically derived rate, and 95% confidence intervals attain stable coverage of 95% ± 0.3%.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Metareasoning and Metaheuristics
• Multiagent Systems: Multiagent Systems under Uncertainty

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Sepsis remains a critical challenge due to its high mortality and complex prognosis. To address data limitations in studying MSSA sepsis, we extend existing transfer learning frameworks to accommodate transformation models for high-dimensional survival data. Specifically, we construct a measurement index based on C-index for intelligently identifying the helpful source datasets, and the target model performance is improved by leveraging information from the identified source datasets via performing the transfer step and debiasing step. We further provide an algorithm to construct confidence intervals for each coefficient component. Another significant development is that statistical properties are rigorously established, including $ell_1/ell_2$-estimation error bounds of the transfer learning algorithm, detection consistency property of the transferable source detection algorithm and asymptotic theories for the confidence interval construction. Extensive simulations and analysis of MIMIC-IV sepsis data demonstrate the estimation and prediction accuracy, and practical advantages of our approach, providing significant improvements in survival estimates for MSSA sepsis patients.

Problem

• Extend transfer learning for high-dimensional survival sepsis data
• Improve target model via intelligent source dataset identification
• Develop algorithm for coefficient confidence intervals and statistical properties

Innovation

• Extends transfer learning for high-dimensional survival data
• Uses C-index to identify helpful source datasets
• Provides algorithm for confidence interval construction