摘要
为了提高经济领域统计数据的预测精度,代数多项式预测模型的建模方法应运而生.该方法使用代数多项式模型拟合给定的经济统计数据,并使用递推最小二乘法(RLS)对多项式拟合模型的加权系数进行递推计算以获得最优模型参数,然后通过获得的最优多项式模型计算未来预测数据.文章以实际统计的经济数据为例进行了仿真计算,研究结果表明,该方法不仅能实现统计数据的高精度拟合,而且具有很好的预测能力,在经济领域具有广阔的应用前景.
In order to improve the prediction accuracy of economic statistical data, modeling method of algebraic polyno- mial prediction model was put forward. In the proposed method, the given economic statistic data was fitted by algebraic poly- nomial model based on recursive least squares （RLS）, in which the optimal weighted coefficients were obtained through recur- sire calculation, and then the future data was computed by the obtained optimal polynomial model. This paper took the actual statistical data as an example to carry on simulation calculation. The research results show that the proposed method can fit the statistical data in high accuracy, and has good prediction ability. Therefore, it will have broad application prospects in the eco- nomic field.
出处
《经济数学》
2014年第1期85-89,共5页
Journal of Quantitative Economics
基金
国家自然科学基金资助项目(61040049)
关键词
多项式模型
递推最小二乘法
数据拟合
预测
polynomial model
recursive least squares （RLS）, data fitting
forecasting
