Scholarlink

This study employs ensemble learning techniques to predict heart conditions using physiological parameters like Electroencephalography (EEG, μV), Electrocardiography (ECG, mV), Electromyography (EMG, mV), Blood Glucose levels (mg/dL), Foot Pressure (kPa), and Body Temperature (°C). Relevant features are extracted and used as independent variables, with a heart condition (status) as the dependent variable. A fivefold cross-validation method evaluates model accuracy, balancing reliability and efficiency. Visualization of accuracy across folds highlights model performance, demonstrating the potential of ensemble learning for health monitoring and diagnosis. Several ensemble learning techniques were used, including Decision Tree (DT) Bagging Classifier, Random Forest (RF) Bagging Classifier, Logistic Regression (LR) Bagging Classifier, Support Vector Machine (SVM) Bagging Classifier, Naïve Bayes (NB) Bagging Classifier, Multi-Level Perceptron (MLP) (Neural Network) Bagging Classifier, K-Nearest Neighbor (KNN) Bagging Classifier, Gradient Boosting Bagging Classifier, Hard Voting Ensemble Model and Soft Voting Ensemble Model. The performance of each model was evaluated based on metrics like the Cross-Validation Score and Mean Cross-Validation (CV) Score. Gradient Boosting Bagging Classifier offers an accuracy of 99.95%, the highest among other techniques.

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

Predictive Modeling for Health Conditions Using Ensemble Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Rishit Mahapatra,
  • Deepak Sethi

摘要

This study employs ensemble learning techniques to predict heart conditions using physiological parameters like Electroencephalography (EEG, μV), Electrocardiography (ECG, mV), Electromyography (EMG, mV), Blood Glucose levels (mg/dL), Foot Pressure (kPa), and Body Temperature (°C). Relevant features are extracted and used as independent variables, with a heart condition (status) as the dependent variable. A fivefold cross-validation method evaluates model accuracy, balancing reliability and efficiency. Visualization of accuracy across folds highlights model performance, demonstrating the potential of ensemble learning for health monitoring and diagnosis. Several ensemble learning techniques were used, including Decision Tree (DT) Bagging Classifier, Random Forest (RF) Bagging Classifier, Logistic Regression (LR) Bagging Classifier, Support Vector Machine (SVM) Bagging Classifier, Naïve Bayes (NB) Bagging Classifier, Multi-Level Perceptron (MLP) (Neural Network) Bagging Classifier, K-Nearest Neighbor (KNN) Bagging Classifier, Gradient Boosting Bagging Classifier, Hard Voting Ensemble Model and Soft Voting Ensemble Model. The performance of each model was evaluated based on metrics like the Cross-Validation Score and Mean Cross-Validation (CV) Score. Gradient Boosting Bagging Classifier offers an accuracy of 99.95%, the highest among other techniques.