Healthy Aging: A Proactive Model to Prevent Self-neglecting Behavior in Smart Homes

Abstract

With a continuous and noticeable shift towards an aging population around the world, a drastic shift in elderly care is required. The older people experience significant decline in physical and mental capacity, which limits the ability to care for themselves. It can cause the self-neglecting behavior that include failing to take medications, neglecting personal hygiene, and not eating well. The research community has already made a shift towards the sensors technology and deep learning techniques to monitor the homes for effective interventions. In this paper, our aim is to further develop this research by developing a proactive model to prevent self-neglecting behavior in aging population. We proposed a deep learning approach, which is based on sequence modeling technique – long short-term memory (LSTM). The experiments are performed on publicly available real smart home dataset, where the residence was living alone. The standard performance metrics are calculated to ensures an acceptable performance for the deployment in the real-world setting. Three case studies are discussed to show the effectiveness of the proactive model to prevent the self-neglecting behavior. It is expected that our model may allow elderly individuals to remain independent in their own homes for longer time and reduce the burden on health care systems.