中国住宅商品房平均销售价格的预测研究

陈睿玉

摘要：本文使用基于IOWA诱导有序加权算术平均算子的变权组合预测模型，基于ARIMA（0，2，2）、指数平滑模型、多元线性回归模型三种单项预测模型构建误差平方和最小的诱导有序加权算术平均组合预测模型。本文首先对其概念及基本原理作简要介绍，再通过对2000-2018年住宅商品房平均销售价格进行拟合预测，并说明其有效性和可行性。结果表明，组合预测模型的平均精度和各项误差度量相对于单项预测而言更好，未来五年的住宅商品房平均销售价格依然呈较快的增长趋势，因此，对住宅商品房销售价格进行调控是非常必要的。

Abstract： In this paper， the variable weight combination forecasting model based on IOWA induced ordered weighted arithmetic mean operator is used， and the least sum of square error induced ordered weighted arithmetic mean combination forecasting model is constructed based on ARIMA （0，2，2）， exponential smoothing model and multiple linear regression model. This paper first introduces its concept and basic principle， and then through the fitting prediction of the average sales price of residential commercial housing in 2000-2018， and explains its effectiveness and feasibility. The results show that the average accuracy and various error measures of the combined forecasting model are better than the single forecasting model， and the average sales price of residential commercial housing in the next five years still shows a rapid growth trend， so it is very necessary to regulate the sales price of residential commercial housing.

關键词：IOWA算子;组合预测模型;住宅商品房平均销售价格