An entropy-based spline approximation technique for the automatic detection of peaks in electrodermal activity data

In this study, we employ a recently developed technique for spline approximation of time series to identify and extract peaks within electrodermal activity data that may be correlated with an increase of stress levels. Our novel approach exploits the statistical meaning of entropy to detect sections of data that deviate significantly from a given baseline signal – a reference peaks free sample that aids in fine-tuning the model's parameters. To illustrate the effectiveness of our methodology, we present results from an analysis of a time series dataset comprising skin conductance measurements collected through a wearable device. The final goal is to develop a tool for monitoring individuals’ stress levels during specific tasks, enhancing their ability to manage stress episodes effectively.