Automatic Prediction of Depression in Older Age

摘要: Maintaining good mental health such as the prevention of severe depressive symptoms is critical for physical and well-being in older adulthood. However, depression elderlies not known quite well thus cannot be treated adequately. In this study, a large wide variety influencing factors from multiple domain areas were investigated using nationally representative sample people English Longitudinal Study Ageing (ELSA). Five different machine learning algorithms employed to build models prediction age. Several model ensemble strategies proposed merge results individual predictive order further improve performance. Significant risk or protective associated with elder separately identified each area. The findings study will enhance our understanding about underlying pathophysiology depression, helping develop appropriate intervention prevent reduce onset