TY - GEN
T1 - On the mathematical modeling of the effect of treatment on human physical activity
AU - Ashour, Mahmoud
AU - Bekiroglu, Korkut
AU - Yang, Chih Hsiang
AU - Lagoa, Constantino Manuel
AU - Conroy, David
AU - Smyth, Joshua Morrison
AU - Lanza, Stephanie Trea
N1 - Publisher Copyright:
©2016 IEEE
PY - 2016/10/10
Y1 - 2016/10/10
N2 - Understanding the main trends in human behavior is fundamental to developing effective adaptive treatments. Inspired by this insight, this paper presents a mathematical quantification of the change of human behavior following external stimuli. In particular, statistical methods are applied to real physical activity data collected intensively using mobile wearable technologies. We explain the setup of the study conducted with multiple participants. Then, a preprocessing of the collected measurements, required to overcome the hurdles associated with behavioral data, is briefly discussed. Furthermore, we identify a dynamical affine model that approximates humans’ sedentary behavior. The affine model is simple yet insightful. We show results of fitting time-invariant as well as switched models along with a quantification of the prediction errors. Moreover, the effect of various types of treatments on the sedentary behavior of several subjects is investigated. As expected, the results show that people react differently to external stimuli. However, common tendencies are clearly observed. Our findings emphasize the necessity of the application of personalized adaptive intervention. Future research directions are discussed accordingly.
AB - Understanding the main trends in human behavior is fundamental to developing effective adaptive treatments. Inspired by this insight, this paper presents a mathematical quantification of the change of human behavior following external stimuli. In particular, statistical methods are applied to real physical activity data collected intensively using mobile wearable technologies. We explain the setup of the study conducted with multiple participants. Then, a preprocessing of the collected measurements, required to overcome the hurdles associated with behavioral data, is briefly discussed. Furthermore, we identify a dynamical affine model that approximates humans’ sedentary behavior. The affine model is simple yet insightful. We show results of fitting time-invariant as well as switched models along with a quantification of the prediction errors. Moreover, the effect of various types of treatments on the sedentary behavior of several subjects is investigated. As expected, the results show that people react differently to external stimuli. However, common tendencies are clearly observed. Our findings emphasize the necessity of the application of personalized adaptive intervention. Future research directions are discussed accordingly.
UR - http://www.scopus.com/inward/record.url?scp=85077885347&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85077885347&partnerID=8YFLogxK
U2 - 10.1109/CCA.2016.7587951
DO - 10.1109/CCA.2016.7587951
M3 - Conference contribution
AN - SCOPUS:85077885347
T3 - 2016 IEEE Conference on Control Applications, CCA 2016
SP - 1084
EP - 1091
BT - 2016 IEEE Conference on Control Applications, CCA 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2016 IEEE Conference on Control Applications, CCA 2016
Y2 - 19 September 2016 through 22 September 2016
ER -