为满足卫生部对三级医院库存周转率的要求,提出一种基于小波变换和相似性度量的线性改进模型.基于滑动窗口的数据流相似性原理检验原始序列,小波分解后,根据其线性特征分别搭建模型分而治之,小波重构综合各分量的预测值得到终值.仿真结果表明,该模型提高了突变节点处的预测精度,在模式和非模式集中有优秀的拟合效果和精准的预测效果,验证了该模型的有效性.%To meet the requirements of the Ministry of Health on tertiary hospitals inventory turnover,an improved linear model based on wavelet transform and similarity measure was presented.Based on data stream similarity principle of sliding window,the original sequence was verified and divided according to its linear characteristics after wavelet decomposition.The predicted values of each component were reconstructed by wavelet transform to obtain the final value.Simulation results show that the proposed model can improve the prediction precision at the discontinuous nodes,and it has good fitting effects and accurate prediction effects both in the model and the non-model set,which verifies the effectiveness of the model.
展开▼
