Scholarlink

Cardiovascular disease (CVD) remains a major global health issue, requiring accurate risk prediction models for early intervention. While traditional models use established risk factors, this study leverages machine learning to improve predictive accuracy by integrating variables like gender, serum cholesterol, and resting blood pressure. A novel approach is proposed to enhance a baseline CVD risk prediction model with machine learning predictions. The performance of this enhanced model using a hybrid dataset showed superior predictive accuracy over the baseline. Feature importance analysis highlighted the significant contributions of gender, serum cholesterol, and resting blood pressure. Initial results from machine learning algorithms were Random Forest (0.83), Logistic Regression (0.77), Decision Trees (0.77), ANN (0.58), and KNN (0.71). With the hybrid dataset, improved accuracies were seen: Random Forest (0.91), Logistic Regression (0.86), Decision Tree (0.83), ANN (0.76) and KNN (0.83). This research refines CVD risk assessment, leading to personalized interventions and better public health outcomes.

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

Integrating Machine Learning into Cardiovascular Disease Risk Prediction: A Comprehensive Analysis of Cholesterol, Heart Rate, and Gender Impact on Disease Prevalence

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Abdul Rahim,
  • Amit Chhabra,
  • Manya,
  • Sunil K. Singh,
  • Sudhakar Kumar,
  • Hardik Gupta,
  • Karan Sharma

摘要

Cardiovascular disease (CVD) remains a major global health issue, requiring accurate risk prediction models for early intervention. While traditional models use established risk factors, this study leverages machine learning to improve predictive accuracy by integrating variables like gender, serum cholesterol, and resting blood pressure. A novel approach is proposed to enhance a baseline CVD risk prediction model with machine learning predictions. The performance of this enhanced model using a hybrid dataset showed superior predictive accuracy over the baseline. Feature importance analysis highlighted the significant contributions of gender, serum cholesterol, and resting blood pressure. Initial results from machine learning algorithms were Random Forest (0.83), Logistic Regression (0.77), Decision Trees (0.77), ANN (0.58), and KNN (0.71). With the hybrid dataset, improved accuracies were seen: Random Forest (0.91), Logistic Regression (0.86), Decision Tree (0.83), ANN (0.76) and KNN (0.83). This research refines CVD risk assessment, leading to personalized interventions and better public health outcomes.