Achieving Model Completeness for Hierarchally Structured Activities of Daily Life

Being able to recognise everyday activities of daily life provides the opportunity of tracking functional decline among elderly people who suffer from Alzheimer’s disease. This paper describes an approach that has been developed for recognising activities of daily life based on a hierarchal structure of plans. While it is logical to envisage that the most common activities will be modelled within a library of plans, it can be impossible to imagine that the library contains plans for every possible hierarchal activity. In order to generalise the activity recognition capability outside the framework of the core activities constructed to support recognition, decision trees are constructed using a well-known induction algorithm during a train period. The motivation of this work is to allow people with Alzheimer’s disease to have additional years of independent living before the disease reaches a stage where it becomes incurable.