Scholarlink

<p>Early diagnosis of hepatitis is essential for improving patient outcomes and reducing disease severity. This paper presents a hybrid machine learning framework called Fuzzy Logic–K Nearest Neighbor–Hybrid Dingo Particle Swarm Optimization–Multistage Ensemble Model (FL-KNN–HDPSO–MSEM), designed to enable efficient and accurate hepatitis detection. The framework integrates three key stages: (i) Fuzzy Logic–based KNN (FL-KNN) for missing value imputation and preprocessing, (ii) Hybrid Dingo–Particle Swarm Optimization (HDPSO) for optimal feature selection, and (iii) Multistage Ensemble Model (MSEM) for final classification. The proposed framework was evaluated using the Hepatitis and ILPR benchmark datasets. Experimental results demonstrate superior performance, achieving accuracies of 98.99% and 98.7%, respectively, outperforming several state-of-the-art models, including AdaBoost, Random Forest, XGBoost, and Support Vector Machine. By integrating preprocessing, feature optimization, and ensemble-based classification within a unified architecture, the proposed model ensures high reliability, improved diagnostic precision, and enhanced computational efficiency for early hepatitis prediction.</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

A hybrid machine learning framework for early hepatitis detection using enhanced feature selection and ensemble classification

  • 期刊论文

  • 发表日期：

    2026年1月10日

查看全文
  • Saranya Jothi C.,
  • R. RoselinKiruba,
  • Vasumathy,
  • Jude Moses Anto Devakanth J.

摘要

Early diagnosis of hepatitis is essential for improving patient outcomes and reducing disease severity. This paper presents a hybrid machine learning framework called Fuzzy Logic–K Nearest Neighbor–Hybrid Dingo Particle Swarm Optimization–Multistage Ensemble Model (FL-KNN–HDPSO–MSEM), designed to enable efficient and accurate hepatitis detection. The framework integrates three key stages: (i) Fuzzy Logic–based KNN (FL-KNN) for missing value imputation and preprocessing, (ii) Hybrid Dingo–Particle Swarm Optimization (HDPSO) for optimal feature selection, and (iii) Multistage Ensemble Model (MSEM) for final classification. The proposed framework was evaluated using the Hepatitis and ILPR benchmark datasets. Experimental results demonstrate superior performance, achieving accuracies of 98.99% and 98.7%, respectively, outperforming several state-of-the-art models, including AdaBoost, Random Forest, XGBoost, and Support Vector Machine. By integrating preprocessing, feature optimization, and ensemble-based classification within a unified architecture, the proposed model ensures high reliability, improved diagnostic precision, and enhanced computational efficiency for early hepatitis prediction.