Random forest and Bayesian prediction for Hepatitis B virus reactivation

Abstract

This paper established the random forest and Bayesian classification prediction models and aims to find out the risk factors of Hepatitis B virus (HBV) reactivation after the precise radiotherapy in patients with primary liver cancer (PLC). Using the identified risk factors we can provide the reference to the doctor to reduce the incidence of the disease. Firstly, we proposed random forest method to select key features and then establish classification prediction models with the key subset. All the features are sorted according to the importance. We select 5 most important features which would be combined into a brand new feature subset and with the new subset we establish random forest and Bayesian classification prediction model. We find that HBV DNA level, TNM tumor staging, V10, V20, outer margin of radiotherapy is the risk factors of HBV reactivation. The classification accuracy of random forest can be reached to 85.15% by using 5 fold cross validation under 200 decision trees, meanwhile, the accuracy of Bayesian classifier reached to 84.57% by using 10 fold cross validation. The experimental results showed that the random forest can be used to evaluate the importance of variables and select the key features. And also, it is a better method to solve the classification prediction problem of HBV reactivation.