基于PCA-BP的高等教育过度预警模型及其应用研究

Study on PCA-BP Based Alarming Model of Over-higher Education and Its Applications

通过专家访谈和文献查阅的方法,初步建立高等教育过度预警指标集,并运用主成分分析法(PCA)对指标进行逐步筛选,确立高等教育过度预警指标体系,然后,将PCA指标逐步筛选法和BP神经网络相结合,构建PCA-BP预警模型,对我国2002年高等教育发展进行评价,从而得出其发展状况在正常至轻微过度之间的结论。

Expert interview and documentary review methods are adopted to collect indexes for the establishment of the alarming model of over-higher education. The alarming index system of overhigher education has been set up by the use of principal component analysis （PCA） index screen. Combining PCA with BP neural network, we used the PCA-BP based alarming model to appraise the development of domestic higher education in 2002, concluding that it is between the normal and the slightly over-developed.