Scholarlink

This study evaluates machine learning (ML) models, including Gradient Boosting (GB), Decision Tree (DT), Neural Network (NN), Extra Trees (ET), K-Nearest Neighbors (K-NN), and XGBoost, for Chronic Kidney Disease (CKD) prediction. Data preprocessing, Recursive Feature Elimination (RFE), and random oversampling were used to address class imbalance and enhance performance. Models were assessed using precision, accuracy, F1-score, Matthews Correlation Coefficient (MCC), and AUC-ROC with five- and tenfold cross validation. Neural network achieved the highest accuracy and precision of 99.40%, and MCC of 0.9880 during tenfold cross validation. Ensemble models, including extra trees and XGBoost, also performed exceptionally with 99.20% accuracy. The results highlight the potential of ensemble methods and feature selection in improving CKD prediction accuracy, supporting early diagnosis and timely intervention strategies.

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 Prediction of Chronic Kidney Disease Using Ensemble Techniques and Machine Learning Models

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Prince Jain,
  • Anand Joshi

摘要

This study evaluates machine learning (ML) models, including Gradient Boosting (GB), Decision Tree (DT), Neural Network (NN), Extra Trees (ET), K-Nearest Neighbors (K-NN), and XGBoost, for Chronic Kidney Disease (CKD) prediction. Data preprocessing, Recursive Feature Elimination (RFE), and random oversampling were used to address class imbalance and enhance performance. Models were assessed using precision, accuracy, F1-score, Matthews Correlation Coefficient (MCC), and AUC-ROC with five- and tenfold cross validation. Neural network achieved the highest accuracy and precision of 99.40%, and MCC of 0.9880 during tenfold cross validation. Ensemble models, including extra trees and XGBoost, also performed exceptionally with 99.20% accuracy. The results highlight the potential of ensemble methods and feature selection in improving CKD prediction accuracy, supporting early diagnosis and timely intervention strategies.