Type: Master
Status: Available

Mood has a significant impact on how people behave, think, and act. Psychological and social science research highlights that physiological aspects, including mood and stress, vary between individuals. Consequently, a universal stress prediction model is ineffective due to these variations, as what works seamlessly for one person may yield inaccurate or inconsistent results for another. However, creating individual models for each person lacks scalability. To address this, the student will utilize an existing large dataset and apply clustering techniques to identify similarities among users. The aim is to develop a machine learning model optimized for the specific characteristics and behavior patterns within these user groups, providing a balanced and scalable solution for personalized stress prediction across a diverse population.

This project is available as a MSc thesis.

For more information contact: Lidia Alecci, Silvia Santini