Stress detection and management in daily life using wearable sensors

Stress has become an integral part of our modern society. Researchers investigated ways to cope with it to alleviate its negative effects on human health, society and economy. At this point, widespread usage of smartphones, smartwatches and smart wrist-bands raised the question of whether we can detect and alleviate stress with them. Although research has traditionally been conducted in laboratory settings, a set of new studies have recently begun to be conducted in ecological environments with unobtrusive wearable devices. In this thesis, we developed a stress detection system for daily life. Unobtrusive wearable devices were used for physiological data collection. For that purpose, we used heart rate variability (HRV) and electrodermal activity (EDA) signals. Modality specific artifact detection and removal algorithms, feature extraction and advanced machine learning methods were proposed. We tested our system in a laboratory environment, restricted, semi-restricted and unrestricted real-life environments by collecting data in each environment. We proposed different techniques to improve the state of the art in real life environments. We worked on prominent environmentspecific research questions. We further examined different stress alleviation methods including those which can be applied indoors. We also discussed promising techniques, alleviation methods and research challenges for daily life stress management.