| 应用R软件对比两种手足口病发病预测模型的效果 |
| |
| 引用本文： | 陈纯,李铁钢,肖新才,杨智聪.应用R软件对比两种手足口病发病预测模型的效果[J].国际流行病学传染病学杂志,2016(2):101-104. |
| |
| 作者姓名： | 陈纯 李铁钢 肖新才 杨智聪 |
| |
| 作者单位： | 510440,广州市疾病预防控制中心传染病预防控制部 |
| |
| 基金项目： | 广东省科技计划项目(2013B021800030)，广东省医学科研基金项目(B2015100)，广州市应用基础研究专项(2013J4100095)，广州市医药卫生科技项目(20141A011062)Science and Technology Planning Project of Guangdong Province(2013B021800030)，Medical Research Fund of Guangdong Province(B2015100)，Application of Basic Research Projects in Guangzhou(2013J4100095)，Medical and Health Science and Technology Project of Guangzhou(20141A011062) |
| |
| 摘 要： | 目的 使用R软件对手足口病发病时间序列模型进行分析与对比,探索较佳的手足口病发病预测模型.方法 利用广州市2009-2013年手足口病月报告病例数作为建模数据,采用R软件中的Holt-Winter指数匀滑模型和季节性差分自回归移动平均模型(SARIMA)分别预测2014年12个月份的发病情况,将预测结果与2014年实际发病数进行比较.结果 对广州市2014年1-12月手足口病发病数进行预测,并将预测值与实际值进行比较,结果显示:采用Holt-Winter指数匀滑法和SARIMA预测结果的平均相对误差分别为-0.40和0.12.结论 SARIMA为较佳的预测分析模型,预测结果能较好地拟合广州市手足口病发病序列并进行短期的有效预测.
|
| 关 键 词： | 手足口病 R软件 预测模型 |
Comparison of two predicting models for hand-foot-mouth disease incidence using R software |
| |
| Abstract: | Objective To compare two predicting models for hand-foot-mouth disease (HFMD) incidence using R software,and to explore the optimal predicting model.Methods Monthly case numbers of HFMD in Guangzhou from 2009 to 2013 were used as the modeling data.Holt-Winter model and seasonal autoregressive integrated moving average model (SARIMA) in R software were used to forecast the HFMD incidences of 12 months in 2014.And the predicting incidences were compared with the real HFMD incidences in 2014.Results Comparing with the actural data,the mean error of the Holt-Winter model was-0.40,and the mean error of the SARIMA model was 0.12.Conclusions The SARIMA model can better fit the incidence of HFMD in Guangzhou,and can be used in short-term prediction. |
| |
| Keywords: | Hand foot and mouth disease R software Predicting model |
| 本文献已被 万方数据 等数据库收录！ |
