基于图神经网络与迁移学习的流行病例数预测

PREDICTION OF EPIDEMIC CASES NUMBER BASED ON GRAPH NEURAL NETWORK AND TRANSFER LEARNING

预测流行病的病例数对研究流行病学和保障卫生安全至关重要,但现有的研究工作很少考虑到实时移动性数据等因素,这一问题给病例数的预测研究带来了挑战。因此,在图神经网络GNN的基础上提出一种新型计算框架-信息聚合网络IAN,既考虑地区病例数据特征,也考虑地区之间的人口移动性数据特征。为了优化各个国家的前期预测模型,在该框架的基础上加入迁移学习方法TL。在四个欧洲国家数据集上的实验结果表明,IAN以及IAN-TL明显优于传统方法,能够有效地降低预测误差。

Predicting the number of cases of an epidemic is essential for studying epidemiology and ensuring health and safety,but the existing researches work rarely consider factors such as real-time mobility data,which poses a challenge to the prediction of the number of cases.Therefore,based on the graph neural network GNN,a new computing framework-information aggregation network IAN,is proposed,which not only considers the characteristics of regional case data,but also considers the characteristics of population mobility data between various regions.In order to optimize the early prediction model of each country,the transfer learning method TL was added on the basis of this framework.Experimental results on data sets of four European countries show that IAN and IAN-TL are significantly better than traditional methods and can effectively reduce prediction errors.