Abstract
Digital health applications are becoming increasingly important for assessingand monitoring the wellbeing of people suffering from mental health conditionslike depression. A common target of said applications is to predict the resultsof self-assessed Patient-Health-Questionnaires (PHQ), indicating currentsymptom severity of depressive individuals. In this work, we explore thepotential of using actively-collected data to predict and forecast daily PHQ-2scores on a newly-collected longitudinal dataset. We obtain a best MAE of 1.417for daily prediction of PHQ-2 scores, which specifically in the used datasethave a range of 0 to 12, using leave-one-subject-out cross-validation, as wellas a best MAE of 1.914 for forecasting PHQ-2 scores using data from up to thelast 7 days. This illustrates the additive value that can be obtained byincorporating actively-collected data in a depression monitoring application.