Machine learning–based detection of problematic smartphone usage patterns in adults using passive mobile sensing and self-report data

The aim of this study is to develop and explore algorithmic models capable of detecting problematic smartphone usage patterns associated with lower mental health in adults. Using passively collected smartphone usage and sensor data in combination with self-reported psychological measures, the study investigates whether problematic usage patterns can be identified reliably from real-world data. The project follows an exploratory, data-driven approach and is intended to generate hypotheses and methodological insights for future confirmatory research.