Abstract

Excessive alcohol use has many negative health consequences, including higher risk of mortality. In addition, in 2006, the estimated economic cost of excessive drinking was $223.5 billion when the combined costs of lost productivity, health and health care, and crime are considered…. Most of these studies are cross-sectional, rely on retrospective reports51 and cannot address within-person, drinking/mood relationships over time.” In order to better understand the physiological effects of substance craving and abuse, we propose, train and test, a model for perceived craving through the use of physiological sensors and an android’s internal sensors. While the smartphone collects near real-time data of the body’s emotional state, participants will also answer survey questions about their emotional state to better assess subjective factors of substance craving. A smartphone application was developed that collects accelerometer data from an android smartphone and affective Q-sensor and then uses machine learning to differentiate the data.So far the program is able to differentiate between three different classes of accelerometer data, that of walking, standing and climbing stairs with 97% accuracy. This data represents how we will classify other types of data such as states of agitation and calm in the wake of substance craving.