Matching Items (2)
Idiographic Models of Walking Behavior for Personalized mHealth Interventions: Some Novel Approaches
Description
This thesis presents the development of idiographic models (i.e., single subject or N = 1) of walking behavior as a means of facilitating the design of control systems to optimize mobile health (mHealth) interventions for sedentary adults. Model-on-Demand (MoD), an adaptive modeling technique, is demonstrated as an ideal method for modeling nonlinear systems with noise on a simulated continuously stirred tank reactor (CSTR). Comparing MoD to AutoRegressive with eXogenous input (ARX) estimation, MoD outperforms ARX in terms of addressing both nonlinearity and noise in the CSTR system. With the CSTR system as an initial proof of concept, MoD is then used to model individual walking behavior using intervention data from participants of HeartSteps, a walking intervention that studies the effect of within-day suggestions. Given the number of possible measured features from which to design the MoD models, as well as the number of model parameters that influence the model’s performance, optimizing MoD models through exhaustive search is infeasible.
Consequently, a discrete implementation of simultaneous perturbation stochastic approximation (DSPSA) is shown to be an efficient algorithm to find optimal models of walking behavior. Combining MoD with DSPSA, models of walking behavior were developed using participant data from Just Walk, a day-to-day walking intervention; MoD outperformed ARX models on both estimation and validation data. DSPSA was also applied to ARX modeling, highlighting the use of DSPSA to not only search over model parameters and features but also data partitioning, as DSPSA was used to evaluate models under various combinations of estimation and validation data from a single participant’s walking data. Results of this thesis point to ARX with DSPSA as a routine means for dynamic model estimation in large-scale behavioral intervention settings.
ContributorsKha, Rachael T (Author) / Rivera, Daniel E (Thesis advisor) / Deng, Shuguang (Committee member) / Muhich, Christopher (Committee member) / Arizona State University (Publisher)
Created2022
Description
Behavior-driven obesity has become one of the most challenging global epidemics since the 1990s, and is presently associated with the leading causes of death in the U.S. and worldwide, including diabetes, cardiovascular disease, strokes, and some forms of cancer. The use of system identification and control engineering principles in the design of novel and perpetually adaptive behavioral health interventions for promoting physical activity and healthy eating has been the central theme in many recent contributions. However, the absence of experimental studies specifically designed with the purpose of developing control-oriented behavioral models has restricted prior efforts in this domain to the use of hypothetical simulations to demonstrate the potential viability of these interventions. In this dissertation, the use of first-of-a-kind, real-life experimental results to develop dynamic, participant-validated behavioral models essential for the design and evaluation of optimized and adaptive behavioral interventions is examined. Following an intergenerational approach, the first part of this work aims to develop a dynamical systems model of intrauterine fetal growth with the prime goal of predicting infant birth weight, which has been associated with subsequent childhood and adult-onset obesity. The use of longitudinal input-output data from the “Healthy Mom Zone” intervention study has enabled the estimation and validation of this fetoplacental model. The second part establishes a set of data-driven behavioral models founded on Social Cognitive Theory (SCT). The “Just Walk” intervention experiment, developed at Arizona State University using system identification principles, has lent a unique opportunity to estimate and validate both black-box and semiphysical SCT models for predicting physical activity behavior. Further, this dissertation addresses some of the model estimation challenges arising from the limitations of “Just Walk”, including the need for developing nontraditional modeling approaches for short datasets, as well as delivers a new theoretical and algorithmic framework for structured state-space model estimation that can be used in a broader set of application domains. Finally, adaptive closed-loop intervention simulations of participant-validated SCT models from “Just Walk” are presented using a Hybrid Model Predictive Control (HMPC) control law. A simple HMPC controller reconfiguration strategy for designing both single- and multi-phase intervention designs is proposed.
ContributorsFreigoun, Mohammad T (Author) / Raupp, Gregory B (Thesis advisor) / Tsakalis, Konstantinos S (Thesis advisor) / Spanias, Andreas S (Committee member) / Forzani, Erica S (Committee member) / Muhich, Christopher L (Committee member) / Arizona State University (Publisher)
Created2021