Interpretable Neural Temporal Point Processes for Modelling Electronic Health Records

arXiv:2404.08007v1 Announce Type: new
Abstract: Electronic Health Records (EHR) can be represented as temporal sequences that record the events (medical visits) from patients. Neural temporal point process (NTPP) has achieved great success in modeling event sequences that occur in continuous time space. However, due to the black-box nature of neural networks, existing NTPP models fall short in explaining the dependencies between different event types. In this paper, inspired by word2vec and Hawkes process, we propose an interpretable framework inf2vec for …

abstract arxiv box continuous cs.lg ehr electronic electronic health records event events health however medical modeling modelling nature patients process processes records space success temporal type