Ensemble Learning-based Smartbed System for Enhanced Patient Care

Abstract

A growing number of feature learning methods, particularly those based on deep learning, have been investigated to derive useful feature representations from large quantities of data. However, applying each model in real time for various research requirements can be challenging. With the common use of smartphones equipped with sensors, ensemble learning has become an area of interest among researchers. By obtaining knowledge of a patient’s mobility, a wide range of services can be provided. Therefore, in this research work, the authors endeavor to detect a patient’s state using sensors attached to the patient’s smartbed. The authors specifically create an ensemble network for greater precision and improved accuracy. This paper is based on using ensemble learning techniques to determine a patient’s state of mobility, and data are gathered from integrated devices in the smartbed. In this study, the authors use ensemble learning to distinguish between various forms of transit, including sleeping, standing, sitting, walking, and emergency states. The authors propose an ensemble network model based on deep learning to enhance the performance and resolve issues that may arise in a singular network. The characteristics generated by the neural networks are merged and relearned in this model. The data used in the trials are taken from the sensors attached to the patient and their smartbed.