Unsupervised discovery of activities of daily living characterized by their periodicity and variability. (October 2015)

• Record Type: Journal Article
• Authors: Soulas, Julie; Lenca, Philippe; Thépaut, André
• Is Part Of: Engineering applications of artificial intelligence. Volume 45(2015:Sep.)
• Page Start: 90
• Page End: 102
• Publication Date: 2015-10
• Subjects: Activities of daily living -- Regular patterns -- Episodes -- Periodicity -- Variability -- Aging at home; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2015.06.006
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: Habits characterize the activities of elderly people. Monitoring their habits and their ability to carry out the activities of daily living is a great challenge in order to improve aging at home. In particular, the detection of changes in regular behavior may help to detect emerging disorders. The emergence of smart homes and sensor networks allows the non-intrusive collection of data describing the activities in the home. The collected data is indeed an objective source to mine periodic patterns representing the habits of a particular individual. Extended Episode Discovery (xED) algorithm is described and discussed. This algorithm searches for regular patterns, highlighting the periodicity and variability of each discovered pattern. This approach allows a high adaptability to different users and lifestyles. Experiments on six real-life datasets illustrate the interest of xED.