Analysis of Rural Health Prediction Machine Learning Algorithms with LSTM-GAN Architecture for 5G Medical Time-Series Analysis

Abstract

For smart health-care services and apps, 4G and other communication standards are employed in the healthcare industry. These technologies are essential to the development of smart healthcare services in the future. As the health care sector expands, several applications are anticipated to generate enormous amounts of data in various formats and sizes. Such vast and varied data requires specific handling with regard to end-to-end delay, bandwidth, latency, and other factors. Information and communication technology is developing quickly today. The system must be created in such a way that it takes into account the preferences of senior people, who have additional needs in terms of their living arrangements and the environment in which they live. In order to examine the most important criteria (feature) needed to develop a model for spotting strange behavior in elderly people, this study recommended using the "Decision Making Trial and Evaluation Laboratory" (DEMATEL). After DEMATEL analyzed the main criteria for predicting odd behavior in the elderly, Convolutional Neural Networks (CNNs) and Long Short-Term Memories (LSTMs) were adopted in the detection of strange behavior in the elderly. The study developed a theory by relating the SIMADL dataset to the behavior of elderly persons and then carried out an experimental investigation with CNN and LSTM-GAN. According to performance assessments, the LSTM has a 97% accuracy rate when it comes to identifying odd behavior in elderly people.