Type: Bachelor Master
Status: Draft
Student:

Advances in wearable technologies have made possible the ubiquitous sensing of physiological signals, like Blood Volume Pressure (BVP) or ElectroDermal Activity (EDA). However, some signals, like EDA, can change the value that is recorded depending on which side of the body the sensor is placed on. The medical literature has explored extensively this phenomenon, called Lateralization, in both EDA and other physiological recordings, but it is not clear the impact it might have on wearable-based applications. In this thesis, the objective is thus to analyse the lateralization effect from physiological signals recorded while people are sleeping. As such, the first a data collection will be run, to integrate some existing dataset. Then, the student will analyse the raw signal as well as extract some hand-crafted features. A Machine Learning task will be then investigated, to gauge the impact of lateralization, e.g., higher or lower performance of the classifier depending on which side the data is trained on.

This project is available as a MSc thesis or as a BSc thesis (with reduced tasks).

For more information contact: Leonardo Alchieri