Chen Yili, He Haoming, Liu Dakang, Zhang Xie, Wang Jingpei, Yang Yixiao
School of Automation and Key Laboratory of Intelligent Information Processing and System Integration of IoT (GDUT), Ministry of Education Guangdong University of Technology Guangzhou China.
111 Center for Intelligent Batch Manufacturing Based on IoT Technology (GDUT) Guangdong University of Technology Guangzhou China.
Optim Control Appl Methods. 2021 Oct 21. doi: 10.1002/oca.2806.
Novel coronavirus pneumonia (COVID-19) epidemic outbreak at the end of 2019 and threaten global public health, social stability, and economic development, which is characterized by highly contagious and asymptomatic infections. At present, governments around the world are taking decisive action to limit the human and economic impact of COVID-19, but very few interventions have been made to target the transmission of asymptomatic infected individuals. Thus, it is a quite crucial and complex problem to make accurate forecasts of epidemic trends, which many types of research dedicated to deal with it. In this article, we set up a novel COVID-19 transmission model by introducing traditional SEIR (susceptible-exposed-infected-removed) disease transmission models into complex network and propose an effective prediction algorithm based on the traditional machine learning algorithm TrustRank, which can predict asymptomatic infected individuals in a population contact network. Our simulation results show that our method largely outperforms the graph neural network algorithm for new coronary pneumonia prediction and our method is also robust and gives good results even if the network information is incomplete.
新型冠状病毒肺炎（COVID-19）于2019年底爆发，威胁着全球公共卫生、社会稳定和经济发展，其特点是具有高度传染性和无症状感染。目前，世界各国政府正在采取果断行动，以限制COVID-19对人类和经济的影响，但针对无症状感染者传播的干预措施却很少。因此，准确预测疫情趋势是一个至关重要且复杂的问题，许多类型的研究都致力于解决这一问题。在本文中，我们通过将传统的SEIR（易感-暴露-感染-康复）疾病传播模型引入复杂网络，建立了一种新型的COVID-19传播模型，并基于传统机器学习算法TrustRank提出了一种有效的预测算法，该算法可以在人群接触网络中预测无症状感染者。我们的模拟结果表明，我们的方法在预测新型冠状病毒肺炎方面大大优于图神经网络算法，并且即使网络信息不完整，我们的方法也具有鲁棒性并能给出良好的结果。
