Suppr超能文献

基于 SARIMA-BPNN 模型预测中国海南省报告的乙型肝炎月发病率。

Prediction of reported monthly incidence of hepatitis B in Hainan Province of China based on SARIMA-BPNN model.

机构信息

Department of Mathematical Statistics, International School of Public Health and One Health, Hainan Medical University, Haikou, China.

Institute of Infectious Disease Prevention and Control, Hainan Center for Disease Control & Prevention, Haikou, Hainan, China.

出版信息

Medicine (Baltimore). 2023 Oct 13;102(41):e35054. doi: 10.1097/MD.0000000000035054.

Abstract

In recent years, the incidence of hepatitis B has been serious in Hainan Province of China. To construct a statistical model of the monthly incidence of hepatitis B in Hainan Province of China and predict the monthly incidence of hepatitis B in 2022. Simple central moving average method and seasonal index were used to analyze the trend and seasonal effects of monthly incidence of hepatitis B. Based on the time series of reported monthly incidence of hepatitis B in Hainan Province from 2017 to 2020, a multiplicative seasonal model (SARIMA), multiplicative seasonal model combined with error back propagation neural network model (SARIMA-BPNN), and a gray prediction model were constructed to fit the incidence, and the time series of monthly incidence of hepatitis B in 2021 was used to verify the accuracy of models. The lowest and highest monthly incidence of hepatitis B in Hainan Province were in February and August, respectively, and MAPE of SARIMA, SARIMA-BPNN, and gray prediction models were 0.089, 0.087, and 0.316, respectively. The best fitting model is the SARIMA-BPNN model. The predicted monthly incidence of hepatitis B in 2022 showed a downward trend, with the steepest decline in March, which indicates that the prevention and control of hepatitis B in Hainan Province is effective, and the study can provide scientific and reasonable suggestions for the prevention and control of hepatitis B in Hainan.

摘要

近年来,中国海南省乙型肝炎发病率较高。为构建中国海南省乙型肝炎月发病率的统计模型,并预测 2022 年乙型肝炎的月发病率。采用简单移动平均法和季节指数法分析乙型肝炎月发病率的趋势和季节性影响。基于 2017 年至 2020 年中国海南省乙型肝炎报告的月发病率时间序列,建立了乘积季节性模型(SARIMA)、乘积季节性模型结合误差反向传播神经网络模型(SARIMA-BPNN)和灰色预测模型来拟合发病率,并利用 2021 年乙型肝炎月发病率时间序列对模型的准确性进行验证。海南省乙型肝炎的最低和最高月发病率分别出现在 2 月和 8 月,SARIMA、SARIMA-BPNN 和灰色预测模型的平均绝对百分比误差(MAPE)分别为 0.089、0.087 和 0.316。最佳拟合模型是 SARIMA-BPNN 模型。2022 年乙型肝炎月发病率的预测呈下降趋势,3 月下降幅度最大,表明海南省乙型肝炎防控效果显著,本研究可为海南省乙型肝炎防控提供科学合理的建议。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f0b9/10578744/be3465eeb764/medi-102-e35054-g001.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验