Simon Perneel
Data scientist
Imec-mict-UGent'
I am a researcher at Imec-mict-UGent, focusing on analyzing smartphone logdata to predict moods and stress levels from smartphone use patterns. I am responsible for the MobileDNA-project, a research tool used to get insights in smartphone use.
Education
KU Leuven
Master Artificial Intelligence
Thesis: Identification of features for the optimal recognition of physical activities with wearable devices
KU Leuven, Campus Ghent
Master Industrial Engineering: Electronics-ICT
Thesis: Examination of Low-Power Wide-Area networks for IoT applications
Experience
Data engineer/scientist — Imec-mict-UGent
Looking for patterns in people's smartphone behavior and use of apps. Investigating whether smartphone use can be a predictor for ones (digital) wellbeing (stress, headaches, moods, online vigilance). Responsible for the MobileDNA research tool. MobileDNA is an app developed by imec-mict-UGent to monitor people's smartphone usage.
Portfolio
Optimizing human activity recogntion with inertial sensors
For my master's thesis, I developed a human activity recognition system using wearable Magnetic Inertial Measurement Units (MIMUs). I compared feature-based and 'raw' inertial data-based classification approaches, evaluating sensor placement, feature selection, and the value of magnetometer data. The final model, using only accelerometer and gyroscope features from 3 optimally placed sensors, achieved an accuracy of 97% and F1-score of 98%.