Scholarlink

<p>Missing data in electronic health records poses a serious challenge. It leads to significant information loss, and traditional methods such as mean, mode, and median introduce bias in the result. This work proposes a novel combination imputation method called Fuzzy Marine Adaptive Clustering (FMAC), which combines optimization based on the Marine Predators Algorithm with fuzzy clustering. It presents a dynamic neighbor selection approach that uses membership score and weighted similarity to handle diverse healthcare data in an accurate way. This multi-step procedure yielded statistically more accurate and better performance than other existing methods. Experiments are done on benchmark healthcare datasets (Diabetes, Parkinson’s, Fetal Heart and H1N1 flu vaccine Disease) to evaluate the effectiveness of the proposed method. The proposed method significantly improves Mean Absolute Error, Mean Squared Error, Root Mean Squared Error, Normalized Root Mean Squared Error, Davies–Bouldin Index, Average Silhouette Coefficient and Friedman and Nemenyi Hypothesis Test. Simulation results show that the proposed algorithm performs better than existing methods. Due to its use of iterative optimization and hybrid similarity computation over large datasets, FMAC benefits from high-performance computing (HPC) resources for real-time imputation and scalability. By enhancing data quality, FMAC supports improved clinical decision-making and predictive analytics in healthcare.</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

Fuzzy marine adaptive clustering with weighted similarity for accurate missing data imputation in electronic health records

  • 期刊论文

  • 发表日期：

    2025年8月22日

查看全文
  • Subhashish Nayak,
  • P. M. Khilar

摘要

Missing data in electronic health records poses a serious challenge. It leads to significant information loss, and traditional methods such as mean, mode, and median introduce bias in the result. This work proposes a novel combination imputation method called Fuzzy Marine Adaptive Clustering (FMAC), which combines optimization based on the Marine Predators Algorithm with fuzzy clustering. It presents a dynamic neighbor selection approach that uses membership score and weighted similarity to handle diverse healthcare data in an accurate way. This multi-step procedure yielded statistically more accurate and better performance than other existing methods. Experiments are done on benchmark healthcare datasets (Diabetes, Parkinson’s, Fetal Heart and H1N1 flu vaccine Disease) to evaluate the effectiveness of the proposed method. The proposed method significantly improves Mean Absolute Error, Mean Squared Error, Root Mean Squared Error, Normalized Root Mean Squared Error, Davies–Bouldin Index, Average Silhouette Coefficient and Friedman and Nemenyi Hypothesis Test. Simulation results show that the proposed algorithm performs better than existing methods. Due to its use of iterative optimization and hybrid similarity computation over large datasets, FMAC benefits from high-performance computing (HPC) resources for real-time imputation and scalability. By enhancing data quality, FMAC supports improved clinical decision-making and predictive analytics in healthcare.