论文部分内容阅读
目的建立临床血液需求量预测的时间序列ARIMA模型,预测临床血液需求量,为血液招募工作提供理论依据。方法收集2006～2010年成都市临床血液月用血量,建立ARIMA模型,预测2011年1～3月需求量。结果临床血液需求量构建为MA(1),1～60月资料所建立模型模型可用作预测,2011年1～3月预测值与实际值基本一致,相对误差依次为1.3%、3.6%、1.5%。结论 采用时间序列MA(1)模型预测短期临床血液需求量,能为开展血液招募采集工作提供科学依据。
Objective To establish a time series ARIMA model for predicting the clinical blood requirement and predict the clinical blood demand so as to provide a theoretical basis for blood recruitment. Methods The monthly blood of clinical blood in Chengdu was collected from 2006 to 2010. The ARIMA model was established to predict the demand from January to March of 2011. Results The clinical blood demand was constructed as MA (1). The model established from January to May was used as a prediction. From January to March in 2011, the predicted and actual values were basically the same with relative errors of 1.3% and 3.6%, respectively. 1.5%. Conclusion Prediction of short-term clinical blood demand using time series MA (1) model can provide a scientific basis for blood recruitment and collection.