Analysing Effectiveness of Multi-Label Classification in Clinical Coding
Rajvir Kaur, Jeewani Anupama Ginige · Proceedings of the Australasian Computer Science Week Multiconference · 2019
In Australia, hospital discharge summaries created at the end of an episode of care contain the patient's medical information based on which clinical codes are assigned. A patient can have multiple diseases and interventions carried out during their stay in the hospital. In this paper, we have done multi-label diseases and interventions classification using Binary Relevance, Label Power-set, and Multi-Layer k-Nearest Neighbor classifier. Our experimental work is divided into three tasks: Random Selection, User Selected, and Repetitive Task. Repetitive task gave better performance in comparison to the other two task.