几种预测方法在甘肃省梅毒发病率预测中的应用

发布时间：2018-10-21 15:49

【摘要】：目的比较几种传统模型及机器学习方法,在甘肃省预测梅毒发病率的效果,并对未来发病率进行预测,为制定控制措施提供依据。方法应用MATLAB 2014a软件,对甘肃省2004-2015年梅毒发病率数据分别建立多项式回归、平滑样条插值、灰色系统GM(1,1)、自回归整合移动平均(ARIMA)、人工神经网络(ANN)和支持向量机(SVR)等数学模型,然后根据2016年实际发病率数据来检验预测效果以选择最佳预测模型,最后使用该模型预测2017-2020年发病率。结果构建的一次多项式、二次多项式、平滑样条方法、GM(1,1)、ARIMA、ANN和SVR模型,拟合2004-2015年梅毒发病率平均相对误差分别为20.04%、22.44%、8.10%、24.89%、11.00%、17.61%和24.72%,以平滑样条最小。7种模型预测2016年梅毒发病率,以ARIMA模型最佳,使用该模型预测2017-2020年发病率分别为19.11/10万、18.21/10万、18.57/10万和19.94/10万。结论不同数学模型拟合和预测效果不同,应根据实际数据选择合适的模型;ARIMA模型预测甘肃省近年梅毒发病率性能较好,预测2017-2020年发病率较为稳定。
[Abstract]:Objective to compare several traditional models and machine learning methods to predict the incidence of syphilis in Gansu Province and to predict the future incidence of syphilis. Methods using MATLAB 2014a software, the mathematical models of syphilis incidence data from 2004 to 2015 in Gansu Province were established, such as polynomial regression, smooth spline interpolation, grey system GM (1 / 1), autoregressive integrated moving average (ARIMA), artificial neural network (ANN) and support vector machine (SVR). Then, according to the actual incidence data of 2016, the prediction effect is tested to select the best prediction model. Finally, the model is used to predict the incidence in 2017-2020. Results the first order polynomial, quadratic polynomial, smoothing spline method, GM (1Q), ARIMA,ANN and SVR models were used to fit the average relative error of syphilis incidence from 2004 to 2015. The average relative errors of syphilis incidence were 20.04 and 22.448.10g, respectively. The incidence rate of syphilis in 2016 was predicted by the smoothing spline model with the minimum of 17.61% and 24.72%, respectively. The ARIMA model is the best one, which is used to predict the incidence of the disease in 2017-2020 at 19.11 / 100,000, 18.21 / 100, 185.7 / 100, and 199,400 / 100, respectively. Conclusion the fitting and forecasting effects of different mathematical models are different, the suitable models should be selected according to the actual data, and the ARIMA model has better performance in predicting syphilis incidence in Gansu Province in recent years, and the incidence rate in 2017-2020 is relatively stable.
【作者单位】：甘肃省疾病预防控制中心;
【基金】：甘肃省卫生行业科研计划资助项目(GSWSKY-2014-22)
【分类号】：R759.1

【相似文献】