基于战略转换的企业战略风险预警模型研究

Study on the Alarm Model of Strategic Risk Based on Strategic Transformation

将粗糙集理论与神经网络相结合,构建了基于战略转换的企业战略风险预警模型。该模型首先运用基于信息熵与MDV函数的模糊聚类算法进行连续属性离散化,然后采用粗糙集理论约简出重要指标体系,最后采用BP神经网络进行学习和训练,进而对检验样本的风险等级进行判断。实证分析表明:基于MDV函数与信息熵的模糊聚类算法能够有效改善离散化效果;添加动量因子的改进BP算法提高了网络学习效率,且该预警模型对检验样本的总体预测精度较高,是一种有效和实用的战略风险预警工具。

In this paper, on the basis of integration of rough sets and neural network a alarm model of strategic risk based on strategic transformation. Firstly, the continuous attribute values are discretized using fuzzy clustering algorithm based on MDV search method and information entropy. And then the attributes are reduced by rough sets. At last, the BP neural network is trained with training samples and the model is used to identify the life cycle of testing samples. The analysis results show that the fuzzy clustering algorithm based on MDV and information entropy can improve the discretization performance effectively. The learning efficiency of BP algorithm is improved by the momentum accession, and the model is an efficient and practical alarm tool of strategic risk.