论文部分内容阅读
给出的基于分数维数的推断方法，用于推断两列观测数据是否来自于同一非线性动态系统，以解决非线性经济预测中的变量选择问题；随后对利用ＢＤＳ检验推断非线性可预测性的方法进行了讨论．在一个决策支持系统中，用该方法对国际有色金属市场价格趋势非线性特征进行统计推断与辅助建模，取得了较好效果．
The proposed method based on fractal dimension is used to infer whether two columns of observational data come from the same nonlinear dynamic system to solve the problem of variable selection in nonlinear economic forecasting. The method is discussed. In a decision support system, this method has been used to make statistical inference and auxiliary modeling on the non-linear trend of international non-ferrous metals market prices, and achieved good results.