Predicting Prevalence of Respiratory Disease with Multi-task Gaussian Process: A Case Study in East China

Abstract

Chronic respiratory disease has become the third leading cause of non-communicable disease death in China. Predicting its prevalence could provide guidance for medical resource allocation given the unbalanced socioeconomic development cross areas. In this study, we investigated the prevalence of respiratory disease in Funan, a representative poverty-stricken county in East China, and developed a multi-task Gaussian process model to predict respiratory disease encounters in Funan. The mean squared error of our model improved from 0.019 to 0.018 compare to using single-task Gaussian process.