档案利用需求的灰色预测方法适用性对比研究

Comparative Study on the Applicability of Gray Prediction Method for Archival Utilization

档案利用需求预测是档案管理的重要环节之一。根据档案利用的时序历史统计数据,利用GM(1,1)模型、DGM(1,1)模型和OLS模型构建了GM-DGM-OLS档案利用需求预测模型,并选取预测有效度方法测算了各模型的权重。采用重庆市长寿区档案馆2014年档案利用数据作为研究样本,分别将各单一预测模型的结果进行了对比分析,同时测度出组合模型的预测结果。对比发现,DGM(1,1)模型在档案利用需求预测中呈现出高精度的特征,而组合预测模型的精度虽然弱于DGM(1,1)模型,但要高于GM(1,1)与OLS模型。因此,在档案利用需求预测实际运用中,建议选用DGM(1,1)与GM-DGM-OLS模型。

Demand prediction is an important part of archive management. Hence, the GM-DGM-OLS forecasting model for archival utilization was constructed based on the model of GM（1,1）, DGM（1,1） and OLS, and then the weight of single model was calculated by the predictive validity method. Moreover, the archival utilization data of Changshou district archives in 2014 was used as a sample of research,and the prediction effect of each model was compared. The results showed that the DGM（1,1） model presented a high-precision feature in the file utilization predict, and the accuracy of combined prediction model was weaker than the DGM（1,1） model, but it was better than GM（1,1） and OLS model. So the DGM（1,1） or GM-DGM-OLS model could be selected to predicting the demand of archival utilization.