As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
This paper discusses advances into the use of anxiety to detect abnormality in a smart home that cares for the occupant. An anxious model of activity has been proposed previously that statistically describes interactions between an occupant and various appliances raising alarms if unusual durations between interactions are detected. To achieve this it models normality only, as abnormality can take too many forms to be easily modelled. This paper extends this work by exploring how the system should learn the statistical parameters for each user and what statistical models are appropriate. Learning is essential to allow the system to be easily integrated into an occupant's lifestyle and recognises that each occupant has different patterns of behaviour. Batch and incremental learning methods and diifferent statistical models are explored using real data for long periods of time. Results show that an incremental learning strategy is most suitable and a Rayleigh distribution is a good choice for the statistical models.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.