Scholarlink

This investigation investigates the application of Slope Boosting Machines (GBM) in foreseeing well-being dangers inside Internet of Things (IoT) based situations. Leveraging differing information collected from IoT gadgets, counting physiological signals, statistical data, natural components, and way of life behaviors, we created prescient models to recognize people at the lifted hazards of different well-being conditions. Through comprehensive experimentation and comparison with conventional machine learning calculations, counting Random Forest and Decision Trees, our comes about illustrate that GBM and its variations, XGBoost and LightGBM, reliably beat in terms of exactness, accuracy, review, and region beneath the ROC curve (AUC). Specifically, LightGBM displayed the most elevated precision of 0.89, exactness of 0.92, review of 0.88, and AUC of 0.94 among all calculations tried. These discoveries emphasize the adequacy of GBM-based approaches in empowering proactive healthcare intercessions and making strides quiet results. Moreover, our inquiry contributes to progressing information in IoT-enabled prescient analytics and underscores the transformative potential of coordination IoT gadgets with advanced machine learning 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

Health Risk Prediction in IoT-Based Environments Using Gradient Boosting Machines

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Pratikkumar Chunawala,
  • Suman Kumar Swarnkar,
  • Aakansha Soy,
  • Omprakash Dewangan,
  • Anupa Sinha,
  • Kamlesh Kumar Yadav,
  • Ankita Singh Baghel

摘要

This investigation investigates the application of Slope Boosting Machines (GBM) in foreseeing well-being dangers inside Internet of Things (IoT) based situations. Leveraging differing information collected from IoT gadgets, counting physiological signals, statistical data, natural components, and way of life behaviors, we created prescient models to recognize people at the lifted hazards of different well-being conditions. Through comprehensive experimentation and comparison with conventional machine learning calculations, counting Random Forest and Decision Trees, our comes about illustrate that GBM and its variations, XGBoost and LightGBM, reliably beat in terms of exactness, accuracy, review, and region beneath the ROC curve (AUC). Specifically, LightGBM displayed the most elevated precision of 0.89, exactness of 0.92, review of 0.88, and AUC of 0.94 among all calculations tried. These discoveries emphasize the adequacy of GBM-based approaches in empowering proactive healthcare intercessions and making strides quiet results. Moreover, our inquiry contributes to progressing information in IoT-enabled prescient analytics and underscores the transformative potential of coordination IoT gadgets with advanced machine learning strategies.