Estimated Impact of COVID-19 on AIDS Occurrence in Mainland China: Based on a Stacked Ensemble Forecasting Model
SSRN Electronic Journal(2022)
摘要
Background and objective: Acquired immunodeficiency syndrome (AIDS) is a public health issue in mainland China. The search query function of the Internet can help people quickly obtain the relevant information needed at the beginning of the disease, which may make the data of the Internet search engine a powerful supplement to the AIDS monitoring work. We studied the relationship between the monthly reported AIDS cases and Baidu Search Index (BSI) in mainland China, and used it for AIDS epidemic prediction. In this paper, we compared the accuracy of single-layer machine learning models and a stacked ensemble model in predicting AIDS cases, and explore the influence of the COVID-19 pandemic on AIDS epidemic in mainland China in 2020.Methods: The Random Forest (RF), Gradient Boosting Machine (GBM) and back-propagation artificial neural network (BP-ANN) models were respectively conducted simulations using AIDS cases data and BSI data in mainland China from 2011 to 2020. And deep learning was used to stack the above three base learner models.Results: The results showed that there was a great relationship between BSI of specific keywords and reported AIDS cases for both increasing and decreasing AIDS epidemic. And model stacking can yield a better prediction result than using a single base learning model. According to the predictive values of our model, the actual number of AIDS cases reported in 2020 was 11,154 fewer than predicted, and the percentage decreased by 15.010%.Conclusions: The number of reported AIDS cases in mainland China has decreased to some extent due to the impact of COVID-19. The stacked ensemble model is more appropriate for predicting human AIDS epidemic trends in mainland China. Therefore, the stacking ensemble algorithm may be of importance for infectious disease prediction.
更多查看译文
关键词
aids occurrence,stacked ensemble forecasting model,mainland china
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要