Random Forest and Sequential Model for Anomalies Detection in the Activities of the People with Dementia

Abstract

Dementia is a very complex disease that affects the ability to think and remember. The thinking ability affects the daily living activities (DLAs) pattern of the people with dementia (PwD) and thus the caregivers and the healthcare professionals that are responsible for assisting PwD can take preventive actions when anomalies in DLAs are identified to improve the health condition of PwD. The main contributions of the article are: (1) the development of an approach for the anomalies detection (AD) in the DLAs of the PwD, (2) the description of the preprocessing of the DLAs data, (3) the presentation of an approach for the detection of the baseline of PwD using Random Forest (RF), (4) the presentation of an approach based on a Sequential Model (SM) for the detection of the baseline of PwD and (5) the AD in DLAs of PwD using the predicted baseline and the data monitored in a day.