CLINICAL DATA SCIENCE PROJECT | Multimodal Fusion of Clinical Data for Suicide Risk Stratification
Master’s capstone | De-identified EHR and clinical-note data • Built a machine-learning pipeline using structured EHR data and unstructured clinical notes to predict 90-day suicide risk; applied temporal train-test splits and evaluated predictive performance. • Prepared analytical outputs to support interpretation of risk patterns and communicate potential operational value for behavioral health workflows.