Published Papers 📚

Electrocardiogram-Based Mental Stress Detection Amid Everyday Activities Using Machine Learning: Model Development and Validation Study

Buelent Uendes, Alex Antonides, Sjors van de Ven, Denise Johanna van der Mee, Eco de Geus, and Mark Hoogendoorn
· J Med Internet Res (fourthcoming) · Paper · Code

We evaluate ML-based mental stress detection using 127 participants across multiple conditions, testing (1) discrimination against a composite no-stress baseline, (2) generalization to unseen stressors and participants, and (3) robustness under reduced sampling rates and number of features for wearable settings.

Start Smart: Leveraging Gradients For Enhancing Mask-based XAI Methods

Buelent Uendes, Shujian Yu, Mark Hoogendoorn
· ICLR 2025 · Paper · Code

A novel gradient-based initialization technique for mask-based XAI methods called StartGrad that improves optimization speed and performance across various state-of-the-art approaches.