ARIMA-SVM组合模型在肺结核发病趋势预测中的应用

目的 研究ARIMA-SVM组合模型在肺结核发病趋势预测中的应用。方法 使用海南省2005—2021年肺结核发病数据，以2005—2020年发病数作为训练集，2021年发病数为验证集，建立ARIMA模型、SVM模型与ARIMA-SVM组合模型，并对三种模型拟合和预测效果进行分析与评价。结果 ARIMA-SVM组合模型数据拟合RMSE、MAPE分别为41.38、1.98%，模型预测RMSE、MAPE分别为45.18、4.84%，拟合和预测效果均优于ARIMA模型与SVM模型。结论ARIMA-SVM组合模型预测效果优于单一模型，更适合我国肺结核发病趋势预测，为我国传染病预测预警提供了新思路。

Application of ARIMA-SVM combination model in predicting the incidence trend of pulmonary tuberculosis

Objective To investigate the application of autoregressive integrated moving average model (ARIMA) and support vector machines (SVM) combination model in predicting the incidence trend of pulmonary tuberculosis. Methods Based on the incidence data of pulmonary tuberculosis in Hainan Province from 2005 to 2021, the ARIMA model, SVM model, and ARIMA-SVM combination model were established, taking the number of cases from 2005 to 2020 as the training set and the number of cases in 2021 as the verification set. The fitting and prediction effects of the three models were analyzed and evaluated. Results The root mean square error (RMSE) and mean absolute percentage error (MAPE) combined model were 41.38% and 1.98%, respectively, and the predicted RMSE and MAPE of the model were 45.18% and 4.84%, respectively. The fitting and prediction results were better than ARIMA model and SVM model. Conclusion The prediction effect of ARIMA-SVM combination model is better than that of single model and more suitable for the prediction of the incidence trend of pulmonary tuberculosis in China, which provides new insight for the prediction and early warning of infectious diseases in China.