Automatic event detection in computational physiology signals

The dataset which we used in this thesis comes from an experiment where the Computational Physiology Laboratory at the University of Houston participated and it was funded in part by the Toyota Class Action Settlement Safety Research and Education Program. The experiment examined the effect of stress load in driving performance, where 59 participants drove on a simulator under different conditions involving regular driving, driving with some stress load (cognitive, emotional or sensorimotor) and driving under an unindented acceleration event. This thesis was focused in the construction of a reliable statistical process control method aiming to will detect the time periods where the subjects have increased levels of anxietyusing as raw information the perinasal perspiration signal. The statistical methodology involved among others the use of Exponentially Weighted Moving Average (EWMA), time series and mixed effects modeling. We should also mention that all the plots and statistical analysis performed using the statistical programming language R.

Autorentext

I hold a Bachelors degree in Mathematics and recently I completed my postgraduate studies in the field of Statistics. During my postgraduate studies, I focused mainly onStatistical Learning and Statistical Process Control. Generally, I am a person who enjoys tolearn new things regarding science and its applications in the real world problems.