The goodness of global fitting in the reconstruction scheme

In the reconstruction scheme, the results for predictions from chaotic time series are accurate. This work introduces the chi-square test for goodness-of-fit and the statistic R-squared in this scenario. These new features facilitate the choice between different predictors and improve the predictive capacity of the LinMapTS package.

在重构框架下，混沌时间序列的预测结果具备较高准确度。本研究在此场景下引入了卡方拟合优度检验（chi-square test for goodness-of-fit）与决定系数（R-squared）两类统计指标。这些新增特征有助于在不同预测器间开展选型工作，同时提升了LinMapTS工具包的预测性能。