定比回归法在价格指数预测中的应用

为了解决各类统计方法在预测价格指数时精度普遍不高的问题，利用定比回归的思想，给出价格指数的一种具有较好精度和直接经济背景的预测方法，并在价格异常波动的年份通过修正回归直线的斜率的方式来得到修正的预测值，使人们在各种情况下都可以对价格指数的走向有一个较为明确的了解。这种方法，在预测各类增长指数型的指标时,可作为传统的线性回归法和灰色模型的一种有效的替代方法。

Price Index Predication with Regression Model on the Same Base Period

In this paper, a method of predicting price index with regression model on the same base period was presented, and a modified value of the price index by revising the slope of the regression line when the prices fluctuate abnormally was given. This method has better accuracy and more direct economic background than classical regression models and gray model, and can be used for the prediction of growth rate such as production, population, and so on.