Scholarlink

Purpose <p>In high-stakes medical domains like hemodialysis, models must provide reliable prediction intervals (PIs) that quantify uncertainty, not just point predictions. Standard ensembles are often static. This study proposes and evaluates a novel Uncertainty-Aware Dynamic Weighting Ensemble (UADWE) framework to improve the reliability of blood pressure PIs during hemodialysis.</p> Methods <p>The framework dynamically assigns weights to a pool of base models based on their localized performance. Competence is measured by each model’s historical ability to generate high-quality PIs on similar instances. This adaptive mechanism was validated on a real-world clinical dataset from a hemodialysis center using a rigorous repeated grouped k-fold cross-validation protocol to ensure robust evaluation.</p> Results <p>The proposed UADWE framework achieved a statistically significant improvement in prediction interval quality (p-value &lt;0.001), obtaining an Interval Score of 50.18 with a small but consistent advantage over Simple Averaging (50.50) and a substantially larger margin over Stacking ensembles (scores &gt; 59). Concurrently, the framework achieved an overall Prediction Interval Coverage Probability of 90%, matching the nominal target; stratified analysis revealed conservative over-coverage in the data-dense central blood-pressure range and progressive under-coverage in the low-density tails (&lt;100 and &gt;180&#xa0;mmHg) for all evaluated models including UADWE. Point prediction accuracy remained competitive with the best individual learners (RMSE <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\approx\)</EquationSource> </InlineEquation> 11.7).</p> Conclusion <p>This work introduces a novel dynamic ensemble framework with improved prediction interval quality. We demonstrate that the synergy between a dynamic, uncertainty-aware strategy and a well-curated, diverse model pool enhances reliable interval estimation, though external validation is needed before clinical translation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantifying Predictive Uncertainty in Hemodialysis: A Dynamic Ensemble Selection Framework Based on Interval Quality

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月21日

查看全文
  • Hsuan-Ming Lin,
  • JrJung Lyu

摘要

Purpose

In high-stakes medical domains like hemodialysis, models must provide reliable prediction intervals (PIs) that quantify uncertainty, not just point predictions. Standard ensembles are often static. This study proposes and evaluates a novel Uncertainty-Aware Dynamic Weighting Ensemble (UADWE) framework to improve the reliability of blood pressure PIs during hemodialysis.

Methods

The framework dynamically assigns weights to a pool of base models based on their localized performance. Competence is measured by each model’s historical ability to generate high-quality PIs on similar instances. This adaptive mechanism was validated on a real-world clinical dataset from a hemodialysis center using a rigorous repeated grouped k-fold cross-validation protocol to ensure robust evaluation.

Results

The proposed UADWE framework achieved a statistically significant improvement in prediction interval quality (p-value <0.001), obtaining an Interval Score of 50.18 with a small but consistent advantage over Simple Averaging (50.50) and a substantially larger margin over Stacking ensembles (scores > 59). Concurrently, the framework achieved an overall Prediction Interval Coverage Probability of 90%, matching the nominal target; stratified analysis revealed conservative over-coverage in the data-dense central blood-pressure range and progressive under-coverage in the low-density tails (<100 and >180 mmHg) for all evaluated models including UADWE. Point prediction accuracy remained competitive with the best individual learners (RMSE \(\approx\) 11.7).

Conclusion

This work introduces a novel dynamic ensemble framework with improved prediction interval quality. We demonstrate that the synergy between a dynamic, uncertainty-aware strategy and a well-curated, diverse model pool enhances reliable interval estimation, though external validation is needed before clinical translation.