An IoT-based health monitoring model using Arduino UNO, Principal Component Analysis, and a Support Vector Machine achieved an accuracy of up to 96.80% in predicting diabetes severity.
An IoT-based monitoring system using Arduino UNO and SVM achieved up to 96.80% accuracy in classifying diabetic patient severity.
Health monitoring has become a crucial aspect in this modern era of society. Various traditional approaches being used in this has their own limitations in terms of data collection and data transfer across public networks. With the advancements in IoT for cloud platforms, many of the research is being carried out in this area. In our proposed model, we accumulate the various parameters of the patients such as pulse rate, glucose level, temperature and blood pressure using Ardunio UNO. The collected data is preprocessed to compute the dominating features resulting for diabetics using Principal Component Analysis (PCA). These features are pushed to the public cloud in real time and machine learning models can be used to notify the patients and hospital authorities for the severity of the patient. In our experiment we have used Support Vector Machine as Training Model. The experiments resulted with an accuracy of upto 96.80%.
Kumar et al. (Fri,) conducted a other in Diabetes. IoT-based health monitoring model using Arduino UNO, PCA, and SVM was evaluated on Model accuracy. An IoT-based health monitoring model using Arduino UNO, Principal Component Analysis, and a Support Vector Machine achieved an accuracy of up to 96.80% in predicting diabetes severity.