多变量整体模式累加多层统计模型的建立及其在组织绩效预测上的应用研究

Setting up Accumulated Multilevel Statistical Models of Multivariate Full Model and Its Application in Forecasting Achievements of Associations

为了解决多层的少样本或无规则数据的建模问题,在一般多层统计模型的基础上提出了多变量整体模式的累加多层统计模型。此模型把累加方法的优点(将无规则数据转化成有规则数据)与多层统计模型结合起来,拓展了多层统计模型的适用范围。从其在香蕉组织绩效的分析以及在仅有两个调查数据香蕉组织形式绩效的预测中,可以看出此模型有较强的实用性。

For modeling the multilevel few sample or irregular data, The accumulated multilevel statistical models of multivariate full model was built on ordinary multilevel statistical models. This model combines advantages of Accumulated way（transform irregular data into regular data） with multilevel statistical models, expand applied range of multilevel statistical models. We can draw the conclusion that the model is strong practical from its application in analysis the achievement of banana Association and forecasting the achievement of banana Associations based on two survey data.