The research of the predicting model and the temporal variation rule for bacillary dysentery in DaLian from 1981 to 2010
Objective To analyze the rule of bacillary dysentery infection and to find best model to forecast the incidence of bacillary dysentery in DaLian. Methods The descriptive methods was carried out based on the incidence of bacillary dysentery from 1981 to 2010, and multiple liner regression method, backpropagation neural network model, ARIMA (p,d,q) (P,D,Q) s model was established according to the monthly incidence and the data of meteorological factors during the same period. Results The incidence rate of bacillary dysentery in Dalian declined from 1157.48/lakh to 25.96/lakh, the average decrease velocity was 12.27％; the mean absolute percentage error and mean-square error of ARIMA(l, 0, 0) (1, 1, 0) 12 were minimum, that of multiple liner regression method were maximum, and that of back-propagation neural network model were between them. Conclusion There was decrease trends of incidence of bacillary dysentery in DaLian, and the ARIMA model with average air temperature as independent variables would be chosen as the best model to forecast the incidence of bacillary dysentery in DaLian.
bacillary dysentery temporal variation rule predicting model
An Qingyu Wu Jun Wang Xiaoli Fan Xuesong Yao Wei
国际会议
Second Joint Biostatistics Symposium(第二届生物统计国际研讨会2012)
北京
英文
173-182
2012-07-08（万方平台首次上网日期，不代表论文的发表时间）