Type: Master
Status: Assigned January 2023
Student: Emilio Lingenthal

Wearable devices provide the means for continuously identify users by performing biometric recognition. The highly subject-dependent nature of physiological data (e.g., heart rate, electrodermal activity, skin temperature) makes it possible to identify the user by tracking those traits over time, as it is already done nowadays with the fingerprint or iris recognition.
This project aims to understand to what extent user identification is possible and if different condition can hamper that capability. After performing data exploration on an existing data, provided by the supervisor, the student will develop a machine learning model to identify users from wearable data. In addition, he/she will test the efficiency of common anonymization techniques, such as noise addition and synthetic data generation.

For more information contact: Lidia Alecci