Identifying irregular activity sequences: an application to passive household monitoring

Abstract

Abstract Approximately one in five people will live to see their 100th birthday due to advancements in modern medicine and other factors. Over 65’s constitute 42% of elective admissions and 43% of emergency admissions to hospitals. Increasingly, people are turning to technology to help improve health and care of the elderly. There is mixed evidence of the success of wearables in older populations with a key barrier being adoption. In contrast, passive sensors such as infra-red motion and plug sensors have had more success. These passive sensors give us a sequence of categorical “trigger” events throughout the day. This paper proposes a method for detecting subtle changes in sequences while taking account of the natural day-to-day variability and differing numbers of “trigger” events per day.