Scholarlink

This study applies a data science approach to model hypertension patterns in university students from Manta, Ecuador, a group underrepresented in existing research. The CRISP-DM methodology was combined with machine learning algorithms and interactive visualization to analyze a dataset of 26,083 clinical, demographic, and behavioral records. Random Forest, Gradient Boosting, and CN2 Rule Induction models achieved high accuracy (AUC and F1-score > 98%). The study identified key risk factors, generated explainable rules, and developed a dashboard to facilitate early intervention. This approach demonstrates the potential of explainable AI to improve cardiovascular health in young populations and resource-limited settings.

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

Risk Pattern Analysis and Prediction of Hypertension in Ecuadorian University Students Using Machine Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Roberto Cobeña,
  • Giusepy Calle,
  • Jorge Pincay,
  • Gabriela Velásquez,
  • Kruskaya Narváez

摘要

This study applies a data science approach to model hypertension patterns in university students from Manta, Ecuador, a group underrepresented in existing research. The CRISP-DM methodology was combined with machine learning algorithms and interactive visualization to analyze a dataset of 26,083 clinical, demographic, and behavioral records. Random Forest, Gradient Boosting, and CN2 Rule Induction models achieved high accuracy (AUC and F1-score > 98%). The study identified key risk factors, generated explainable rules, and developed a dashboard to facilitate early intervention. This approach demonstrates the potential of explainable AI to improve cardiovascular health in young populations and resource-limited settings.