我国有色金属出口额的组合预测模型

A COMBINATORIAL PREDICTION MODEL FOR THE EXPORT VALUE OF CHINA'S NONFERROUS METALS

有色金属产品是我国出口商品结构中的一个重要组成部分,在竞争激烈的国际市场上,中国有色金属产品的出口前景如何,需要我们用科学的方法加以预测。本文以一个典型产品(锑品)和一个典型市场(美国市场)为例,分别用二元线性回归模型和时间序列分析,对今后八年我国有色金属产品出口额进行了预测,并在这两种预测方法的基础上,推出了一种预测误差更小的组合预测模型。本文还对组合预测模型中权重的确定和预测误差的无偏性进行了证明。

Nonferrous metal products are an important part in China' s export commodities. In the fiercely eom-petitive world market, what a bright future the export of our nonferrous metals will have should be pre-dicted scientifically. Illustrating with a typical product (i. e. antimony goods) and a typical market (i. e.the market of USA), the paper makes a prediction for the export value of our nonferrous metal products inthe eight years to come by means of the binary linear regression model and the time series analysis. Basedon the information these two prediction methods provide, the paper has deduced a combinatorial predictionmodel with very small error of prediction. In the paper the determination of weights and the unbiassednessof prediction error in the combinatorial prediction model are testified as well.