Scholarlink

<p>This study introduces a transparent and interpretable ensemble-based recommender system for the early diagnosis of Non-Alcoholic Fatty Liver Disease (NAFLD). The proposed MZ_Ensemble model integrates Support Vector Machine, Random Forest, and Gradient Boosting classifiers to improve predictive accuracy. Dual-layer interpretability is achieved using SHAP (for global feature impact) and LIME (for case-specific explanation), enabling clinical transparency. The model is validated on a publicly available liver dataset, achieving 94% accuracy with balanced precision and recall. The system is designed as a modular diagnostic tool, capable of supporting real-time decision-making in clinical environments. This work offers a significant advancement in explainable medical AI by bridging accuracy with interpretability.</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

Precision management of NAFLD using advanced machine learning: enhancing predictive diagnostics and personalized treatment

  • 期刊论文

  • 发表日期：

    2025年9月6日

查看全文
  • Md. Zeyauddin,
  • Shafiqulabidin

摘要

This study introduces a transparent and interpretable ensemble-based recommender system for the early diagnosis of Non-Alcoholic Fatty Liver Disease (NAFLD). The proposed MZ_Ensemble model integrates Support Vector Machine, Random Forest, and Gradient Boosting classifiers to improve predictive accuracy. Dual-layer interpretability is achieved using SHAP (for global feature impact) and LIME (for case-specific explanation), enabling clinical transparency. The model is validated on a publicly available liver dataset, achieving 94% accuracy with balanced precision and recall. The system is designed as a modular diagnostic tool, capable of supporting real-time decision-making in clinical environments. This work offers a significant advancement in explainable medical AI by bridging accuracy with interpretability.