Machine Learning-based Classification of Symptom Clusters and Online CBT

To breakthrough the bottleneck identified, we will conduct a cross-sectional study to develop a symptom clustering model for depression and anxiety. A wide range of statistical methods as well as machine learning approaches were explored, and a cohesive hierarchical clustering algorithm will be used. After developing the model, a symptom-matched intervention program based on problem solving therapy will be formulated. We are supposed to examine whether its use for personalizing symptom-matched psychological treatment can lead to improved patient outcomes, compared with usual care. This project is expected to provide a new and precise method for the emotion management, which will provide a standardized intervention pathway combining screening with treatment for the management of depression symptom and anxiety symptom. A preciser intervention matched to individual symptoms may provide important insight in improving patient outcome as well as a standardized mood management pathway targeting to the early detection and intervention for community residents.

Inclusion Criteria: * Aged between 18 and 64 years. PHQ-9 ≥10 and/or GAD-7 ≥8 at baseline assessment defined as the threshold for caseness. Exclusion Criteria: * People will be excluded if they meet any of the following criteria: 1. They are receiving psychological therapy during an interview for any mental health issue; 2. currently acutely suicidal or have attempted suicide in the past 2 months, as indicated by PHQ-9 item 9; 3. cognitively impaired or diagnosed with bipolar disorder or psychosis or experiencing psychotic symptoms; d) dependent on alcohol or drugs; e) living with an unstable or acute medical illness that would interfere with trial participation.