MC-TRCM: Observation-Aware Recursive Fusion for Incomplete Mobile and Wearable Mental-Health Feature Views
The MC-TRCM model addresses learning from incomplete mobile and wearable mental health data by treating every feature source as a separate token. It incorporates missingness directly into the input context to handle incomplete views. This architecture uses modality gating, task conditioning via FiLM, and recursive decoding to produce accurate predictions across various endpoints.