Automatic assessment of problem behavior in individuals with developmental disabilities (2012)

Author(s): Ploetz T, Hammerla NY, Rozga A, Reavis A, Call N, Abowd GD

    Abstract: Severe behavior problems of children with developmental disabilities often require intervention by specialists. These specialists rely on direct observation of the behavior, usu- ally in a controlled clinical environment. In this paper, we present a technique for using on-body accelerometers to as- sist in automated classification of problem behavior during such direct observation. Using simulated data of episodes of severe behavior acted out by trained specialists, we demon- strate how machine learning techniques can be used to seg- ment relevant behavioral episodes from a continuous sensor stream and to classify them into distinct categories of se- vere behavior (aggression, disruption, and self-injury). We further validate our approach by demonstrating it produces no false positives when applied to a publically accessible dataset of activities of daily living. Finally, we show promis- ing classification results when our sensing and analysis sys- tem is applied to data from a real assessment session con- ducted with a child exhibiting problem behaviors.

      • Date: 5-8 September 2012
      • Conference Name: Ubicomp 2012: Proceedings of the 2012 ACM Conference on Ubiquitous Computing
      • Pages: 391-400
      • Publisher: ACM
      • Publication type: Conference Proceedings (inc. abstract)
      • Bibliographic status: Published
      Staff

      Dr Thomas Ploetz
      Lecturer (Context Aware Computing)