Machine learning can improve prediction of hepatitis B surface antigen seroclearance: A large multicentre cohort study in the United Kingdom

Aims:

• To characterise the characteristics of individuals with HBsAg seroclearance versus non-seroclearance.
• To utilise routinely collected electronic patient records from large multicentre cohorts for predictive modelling of HBsAg seroclearance.
• To identify important factors at disease index that contribute to prediction performance.