A Low-Cost IoT Wearable Device with XGBoost, CNN and SVM for Early Detection of Fever, Tachycardia and Hypoxia

Chronic and cardiovascular diseases present a significant global health threat, underscoring the need for remote monitoring technologies capable of ensuring continuous and accessible care. Vital signs such as body temperature, heart rate, and blood oxygen saturation are critical indicators for early detection of health alterations. This study proposes the design of a low-cost wearable device with non-invasive sensors for real-time acquisition and processing of these variables, integrating machine learning algorithms including Support Vector Machines (SVM), Convolutional Neural Networks (CNN), and XGBoost. A dataset of 2,480 samples (2,130 experimental, 350 public) was used for training and validation. The models achieved high predictive performance, with XGBoost obtaining an R² of 0.9765, accuracy of 95.8%, and F1-score of 0.96, surpassing SVM and CNN. These results highlight the potential of combining affordable wearable devices with advanced ML to enable early detection, preventive monitoring, and scalable healthcare solutions.