Skip to main content
Fig. 1 | BMC Medical Informatics and Decision Making

Fig. 1

From: Development of a data-driven COVID-19 prognostication tool to inform triage and step-down care for hospitalised patients in Hong Kong: a population-based cohort study

Fig. 1

Top 20 Features* ranked according to importance in the XGBoost model. *Feature importance of total protein, gender, pneumonia, chronic disease, sore throat, fatigue, myalgia, chills, headache, and vomiting under Day-1 model and that of total protein, myalgia, pneumonia, fever, vomiting, chronic disease, headache, fatigue, sore throat, chills under Day-5 model were excluded from this figure

Back to article page