Multimodal Attention-based Deep Learning for Emergency Triage with Electronic Health Records
A multimodal deep learning architecture that can effectively handle both tabular and textual data is proposed and the potential of the proposed model in predicting triage decisions is demonstrated.