AI-Powered Personal Digital Twin for Stress and Sleep
Autor(ka) práce:
Pheap, Oudamkosal
Typ práce:
Diploma thesis
Vedoucí práce:
Zimmermann, Pavel
Oponenti práce:
-
Jazyk práce:
English
Abstrakt:
This thesis designs and evaluates an AI-powered Personal Digital Twin framework for predicting stress and sleep quality among university students. The framework combines population-level baseline models with personalized models trained on daily wearable and self-reported data. A 30-day field study with 33 students was used to evaluate whether the system can operate under realistic small-data conditions. The results show that the full PDT pipeline reduces prediction error compared with the population baseline and that model performance depends on data availability and individual differences.
Klíčová slova:
Stress Prediction; Personal Digital Twin; Machine Learning; Wearable Data; Sleep Prediction
Název práce:
AI-Powered Personal Digital Twin for Stress and Sleep
Autor(ka) práce:
Pheap, Oudamkosal
Typ práce:
Diplomová práce
Vedoucí práce:
Zimmermann, Pavel
Oponenti práce:
-
Jazyk práce:
English
Abstrakt:
This thesis designs and evaluates an AI-powered Personal Digital Twin framework for predicting stress and sleep quality among university students. The framework combines population-level baseline models with personalized models trained on daily wearable and self-reported data. A 30-day field study with 33 students was used to evaluate whether the system can operate under realistic small-data conditions. The results show that the full PDT pipeline reduces prediction error compared with the population baseline and that model performance depends on data availability and individual differences.
Klíčová slova:
Machine Learning; Personal Digital Twin; Wearable Data; Sleep Prediction; Stress Prediction