Abnormal Behavior Prediction in Elderly Persons Using Deep Learning

• Published in: 2023 International Conference on Computer, Electronics & Electrical Engineering & their Applications (IC2E3) pp. 1 - 5
• Main Authors: Ramana, Attili Venkata, Bhoga, Ujwala, Dhulipalla, Ramya Krishna, Kiran, Ajmeera, Chary, B Deepthi, Chandra Shaker Reddy, Pundru
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08-06-2023
• Subjects: ADL; Anomaly detection; Behavioral sciences; Deep learning; LSTM; Older adults; Privacy; Recurrent neural networks; Sensors; Smart home; Smart homes
• DOI: 10.1109/IC2E357697.2023.10262547

A "smart home" is one in which conveniences like remote monitoring and helpful services are installed to help their residents maintain their sense of autonomy and privacy. When helping an elderly person, it's important to be able to distinguish between the person's typical and atypical actions. Abnormal behaviours noticed while performing ADLs are an excellent predictor of the presence of health and behavioural issues requiring care and support. In this paper, we offer a method for automatically detecting aberrant behaviours in the elderly using long short-term memory recurrent neural networks (LSTM). Our technique can be used to simulate the time-dependent information contained in longitudinal data sets. The purpose of this research is to assess how well LSTM can detect and foresee inappropriate behaviour in the elderly within connected homes. We conducted extensive experiments on a dataset to show that the suggested strategy is effective in predicting abnormal behaviours with a high degree of precision. We also showed that our method is far better than the current gold standard.