One specific regression model that is used in trade remedy investigations is the Granger-causality model. An economic variable x, say dumping, is said to Granger-cause another variable y, say losses to an industry, if past values of x provides information for predicting current and future values of y. In the context of a vector autoregression (VAR), which is the manner in which Granger-causality is carried out by economists, x is said to Granger-cause y if the addition of past values of x to a regression, involving a range of other explanatory variables to predict future values of y, results in an improvement in the prediction (e.g. a statistically significant reduction in the mean square error).
热心网友
In the context of a vector autoregression (VAR), which is the manner in which Granger-causality is carried out by economists, x is said to Granger-cause y if the addition of past values of x to a regression, involving a range of other explanatory variables to predict future values of y, results in an improvement in the prediction (e。g。 a statistically significant reduction in the mean square error)。 在救济调查中采用的一种特定的回归模型是格兰杰因果模型。这种模型认为，如果一个经济变量X(比如倾销）的过去的数值为另外一个变量y（比如其对某行业带来的损失）提供了预测现值和未来值得相关信息，那么这两个变量x和y之间就存在格兰杰因果关系。在矢量自回归--也就是经济学家们实现格兰杰因果关系的方式--情况下，如果X过去数值相加并回归并涉及其他解释性变量来预测y变量的未来值，从而带来预测结果的改善（例如，统计方面平方数值的错误的显著减少）,那么我们就认为X和Y之间有格兰杰因果关系。参考参考吧。有大量的条件句在里面。出差在外，没办法仔细核对，你看看吧。