Detection of Dementia-Related Abnormal Behaviour Using Recursive Auto-Encoders

Age-related health issues have been increasing with the rise of life expectancy all over world. One these problems is cognitive impairment, which causes elderly people to performing their daily activities. Detection impairment at an early stage would enable medical doctors deepen diagnosis and follow-up on patient status. Recent studies show that activities can be used assess status people. Additionally, intrinsic structure relationships between sub-activities are important clues for capturing abilities seniors. Existing methods perceive each activity as a stand-alone unit while ignoring inner structural relationships. This study investigates such by modelling hierarchically from sub-activities, overall goal detecting abnormal linked impairment. For this purpose, recursive auto-encoders (RAE) linear vs. greedy supervised semi-supervised variants adopted model Then, systematically detected using RAE’s reconstruction error. Moreover, apply RAEs problem, we introduce new sensor representation called raw measurement (RSM) captures activities, frequency order activations. As real-world data not accessible, generated simulating behaviour, reflects Extensive experiments decision-supporting tool, especially when training set labelled detect indicators dementia.