Scholarlink

Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide, disproportionately affecting populations in low- and middle-income countries, including Bangladesh. This research introduces a comprehensive machine learning (ML) and deep learning (DL) framework for the early prediction of heart disease, aiming to improve diagnostic accuracy and interpretability through systematic feature selection. A curated dataset related to heart disease was utilized, employing four established feature selection techniques—Mutual Information, Chi-Square, ANOVA F-score, and Recursive Feature Elimination—to identify the most relevant clinical attributes. The study implemented and rigorously evaluated a ML model, K-Nearest Neighbors (KNN), alongside advanced DL models such as Artificial Neural Networks (ANN), Multilayer Perceptron (MLP), 1D Convolutional Neural Networks (1D-CNN), Long Short-Term Memory (LSTM), and Bidirectional Long Short-Term Memory (Bi-LSTM). These models were benchmarked using standard performance metrics, including accuracy, precision, recall, F1-score, and ROC-AUC. The results demonstrate that DL models, particularly ANN and Bi-LSTM, achieved near-optimal classification performance when coupled with appropriate feature selection techniques. The findings highlight the vital role of tailored feature engineering in enhancing model efficacy and underscore the potential of AI-driven diagnostic tools to advance cardiovascular healthcare delivery, especially in resource-constrained environments.

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

Enhanced Cardiovascular Disease Prediction Using Machine Learning and Deep Learning Models with Optimized Feature Selection Techniques

  • 章节

  • 发表日期：

    2026年

查看全文
  • Kamrul Golder,
  • Md Mahmudul Haque,
  • M. Raihan

摘要

Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide, disproportionately affecting populations in low- and middle-income countries, including Bangladesh. This research introduces a comprehensive machine learning (ML) and deep learning (DL) framework for the early prediction of heart disease, aiming to improve diagnostic accuracy and interpretability through systematic feature selection. A curated dataset related to heart disease was utilized, employing four established feature selection techniques—Mutual Information, Chi-Square, ANOVA F-score, and Recursive Feature Elimination—to identify the most relevant clinical attributes. The study implemented and rigorously evaluated a ML model, K-Nearest Neighbors (KNN), alongside advanced DL models such as Artificial Neural Networks (ANN), Multilayer Perceptron (MLP), 1D Convolutional Neural Networks (1D-CNN), Long Short-Term Memory (LSTM), and Bidirectional Long Short-Term Memory (Bi-LSTM). These models were benchmarked using standard performance metrics, including accuracy, precision, recall, F1-score, and ROC-AUC. The results demonstrate that DL models, particularly ANN and Bi-LSTM, achieved near-optimal classification performance when coupled with appropriate feature selection techniques. The findings highlight the vital role of tailored feature engineering in enhancing model efficacy and underscore the potential of AI-driven diagnostic tools to advance cardiovascular healthcare delivery, especially in resource-constrained environments.