目的基于药品零售价格大数据构建药品价格指数,描述其波动特征,发挥其药品价格宏观监管作用,促进药品价格保持合理水平。方法运用链式拉氏指数构建原理建立药品价格指数模型,运用时间序列模型描述指数波动特征,识别并分析药品价格波动...目的基于药品零售价格大数据构建药品价格指数,描述其波动特征,发挥其药品价格宏观监管作用,促进药品价格保持合理水平。方法运用链式拉氏指数构建原理建立药品价格指数模型,运用时间序列模型描述指数波动特征,识别并分析药品价格波动异常状况。结果2015年1月—2020年12月,药品价格总指数小幅上涨,累计涨幅为14.43%,年均涨幅约2.40%,市场化改革成效较为显著。通过基于局部加权回归的季节趋势分解(seasonal-trend decomposition using loess,STL)方法对获得的药品价格总指数时间序列进行分析,指数呈长期平缓上升趋势,不规则波动值为-1.41~2.03,说明药品价格受外因影响较小,周期性特征仍有待进一步研究。2015年1月—2020年12月,根据药品价格指数共监测到价格异常风险32次。结论药品价格指数较全面地反映药品价格走势,对于药品价格异常波动具有一定的预警作用,能够为我国药品价格监管提供有效工具。展开更多
This research takes the view that the modelling of temporal data is a fundamental step towards the solution of capturing semantics of time. The problemsinhereat in the mod6iling of time are not unique to datahase proc...This research takes the view that the modelling of temporal data is a fundamental step towards the solution of capturing semantics of time. The problemsinhereat in the mod6iling of time are not unique to datahase processing. Therepresentation of temporal knowledge and temporal reasoning arises in a widerange of other disciplines. ln this paper an account is given of a techniquefor modelling the semantics of temporal data and its associated normalizationmcthod. It discusses the techniques of processing temporal data by employinga Time Sequence (TS) data model. It shows a number of different strategieswhich are used to classify different data properties of temporal data, and it goeson.to develop the model of temporal data and addresses issues of temporal dataapplication design by introducing the concept of temporal data normalisation.展开更多
文摘目的基于药品零售价格大数据构建药品价格指数,描述其波动特征,发挥其药品价格宏观监管作用,促进药品价格保持合理水平。方法运用链式拉氏指数构建原理建立药品价格指数模型,运用时间序列模型描述指数波动特征,识别并分析药品价格波动异常状况。结果2015年1月—2020年12月,药品价格总指数小幅上涨,累计涨幅为14.43%,年均涨幅约2.40%,市场化改革成效较为显著。通过基于局部加权回归的季节趋势分解(seasonal-trend decomposition using loess,STL)方法对获得的药品价格总指数时间序列进行分析,指数呈长期平缓上升趋势,不规则波动值为-1.41~2.03,说明药品价格受外因影响较小,周期性特征仍有待进一步研究。2015年1月—2020年12月,根据药品价格指数共监测到价格异常风险32次。结论药品价格指数较全面地反映药品价格走势,对于药品价格异常波动具有一定的预警作用,能够为我国药品价格监管提供有效工具。
文摘This research takes the view that the modelling of temporal data is a fundamental step towards the solution of capturing semantics of time. The problemsinhereat in the mod6iling of time are not unique to datahase processing. Therepresentation of temporal knowledge and temporal reasoning arises in a widerange of other disciplines. ln this paper an account is given of a techniquefor modelling the semantics of temporal data and its associated normalizationmcthod. It discusses the techniques of processing temporal data by employinga Time Sequence (TS) data model. It shows a number of different strategieswhich are used to classify different data properties of temporal data, and it goeson.to develop the model of temporal data and addresses issues of temporal dataapplication design by introducing the concept of temporal data normalisation.
