【摘要】 提出一种基于主成分分析的变维数策略,用以克服自适应神经模糊推理系统(ANFIS)难以适应数据多维数情况的缺陷。然后结合企业财务分析领域的特点,将优化的自适应神经模糊推理系统应用于公司财务状况诊断。最后,利用样本公司实际指标数据对模型的诊断能力进行了实证研究,结果显示了这种方法的优越性。更多还原

【Abstract】 <Abstrcat> In order to overcome the weak discriminant power for adaptive neural-fuzzy inference (system)(ANFIS) when tackling multi-dimensional data, an alterable-dimension strategy based on (primary) component analysis is presented. Then the optimized ANFIS method is used to diagnose (corporate) financial conditions. Finally, experiential research is carried out to verify the diagnostic (power) of the proposed method, which demonstrates the superiority of this method.更多还原