On the Forecasting of Body Temperature using IoT and Machine Learning Techniques

Abstract

Artificial Intelligence (AI) knows a high exploitation in medical computing to enhance patient care by accelerating processes and increasing accuracy, thus providing improvements healthcare in general. Temperature is an important health factor that has to be regularly monitored and even early detected in some situations. Thus, this paper aims to invest in the advances in Internet of Things (IoT) and in Machine Learning (ML) techniques to develop a monitoring system that is able to forecast body temperature. The proposed solution consists in: i) designing and implementing a wearable device using a temperature sensor and a micro-controller, to monitor body temperature permanently, then ii) those monitored measurements are collected and stored as a time-series dataset in a cloud storage server accessible by doctors, and iii) finally the time-series dataset is used by machine learning forecasting techniques to get early body temperature values for the next hours.