Scholarlink

<p>Medical diagnosis is a complicated task since it seeks accuracy and efficiency particularly, for cardiovascular disease (CVD). Earlier diagnosis is significant to prevent premature deaths in a significant way, as well as to reduce complications through providing a timely intervention. The current revolution in medical diagnosis uses medical computational intelligence to predict CVD at the earlier stage, but yielding accurate performance is still a challenging task. The conventional diagnostic models faced several challenges concerning higher computational complexity, missing data, lack of generalization, as well as lower exploratory analysis that limits the models from visualizing the data distributions and relationships between different variables. This research designs a medical diagnosis system using the proposed Alpha Hunting optimizer-enabled Neural Backed Decision Tree (AH-NBDT) model in detecting CVD at an earlier stage. The AH-NBDT model effectively solves the diagnosis issues and provides accurate detection at the beginning stage using the NBDT classifier. It effectively learns higher-level concepts and relies on generating proper decisions that further enable to production of more accurate results. It reduces the error value through the loss function and improves the model accuracy by providing a better convergence rate. Moreover, the extraction of optimal features enables the model to detect the disease more accurately by increasing the training speed. The extraction of features helps to reduce memory utilization and minimize the computational complexity issues that occur during model training. The AH-NBDT model is effective against the conventional approaches and exhibits 96.80% accuracy, 96.09% F1-score, 95.80% precision, and 96.38% recall using the heart disease dataset. The proposed AH-NBDT framework provides accurate diagnostic outcomes in identifying the risk factors strongly associated with CVD. Moreover, scalable diagnosis potentially allows for personalized treatment with efficient screening, thereby promoting early disease diagnosis and management.</p>

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

AH-NBDT: Alpha Hunting Optimizer Enabled Neural-Backed Decision Tree Algorithm for Cardiovascular Disease Detection with Big Data in IoT

  • 期刊论文

  • 发表日期：

    2025年7月22日

查看全文
  • Pravin M. Tambe,
  • Manish Shrivastava

摘要

Medical diagnosis is a complicated task since it seeks accuracy and efficiency particularly, for cardiovascular disease (CVD). Earlier diagnosis is significant to prevent premature deaths in a significant way, as well as to reduce complications through providing a timely intervention. The current revolution in medical diagnosis uses medical computational intelligence to predict CVD at the earlier stage, but yielding accurate performance is still a challenging task. The conventional diagnostic models faced several challenges concerning higher computational complexity, missing data, lack of generalization, as well as lower exploratory analysis that limits the models from visualizing the data distributions and relationships between different variables. This research designs a medical diagnosis system using the proposed Alpha Hunting optimizer-enabled Neural Backed Decision Tree (AH-NBDT) model in detecting CVD at an earlier stage. The AH-NBDT model effectively solves the diagnosis issues and provides accurate detection at the beginning stage using the NBDT classifier. It effectively learns higher-level concepts and relies on generating proper decisions that further enable to production of more accurate results. It reduces the error value through the loss function and improves the model accuracy by providing a better convergence rate. Moreover, the extraction of optimal features enables the model to detect the disease more accurately by increasing the training speed. The extraction of features helps to reduce memory utilization and minimize the computational complexity issues that occur during model training. The AH-NBDT model is effective against the conventional approaches and exhibits 96.80% accuracy, 96.09% F1-score, 95.80% precision, and 96.38% recall using the heart disease dataset. The proposed AH-NBDT framework provides accurate diagnostic outcomes in identifying the risk factors strongly associated with CVD. Moreover, scalable diagnosis potentially allows for personalized treatment with efficient screening, thereby promoting early disease diagnosis and management.