Elderly Behavior Prediction Using a Deep Learning Model in Smart Homes

Abstract

The goal of a smart home is to keep track of the behaviors of the older adults with disabilities within the home, and then anticipate their activity to help with other actions. Elderly and disabled people have problems with their daily lives, while most other people are unaware of their difficulties. Helping the elderly to live independently allows them to lead their daily lives in a better manner. The implementation of analytics and machine learning algorithms leads to a predictive approach to health care services. In this chapter, a learning model in a smart home concept focuses on making it possible for the elderly to remain safe and comfortable at home. The transformative home security device learning architecture of the smart home platform is a valuable solution to studying mobility patterns at home, with the ability to identify behavioral changes related to issues of wellbeing. A predictive learning system can effectively recognize and identify the behavior of the elderly. A learning model, a recurrent neural network (RNN) is proposed to evaluate the people's activity. The focus of the present study is to forecast the deterioration in mental function and give warnings for the benefit of seniors.