基于优化概率神经网络的制造业财务预警研究

Research on manufacturing financial early warning based on optimized PNN

财务预警通过对企业相关指标分析构建出预测模型,达到对其风险进行预测的目的,可为利益相关者的关联决策提供依据,使得预警效率的研究成为重点。以90家制造企业的相关数据构成样本搭建概率神经网络模型进行预警研究,为提升模型的效率,引入粒子群算法对模型进行优化。实证分析中得出,未用粒子群算法优化前模型的预测准确率为87.5%,经优化后模型的预测正确率为93.75%。则使用粒子群算法对神经网络的优化的可行性较高,这可做为财务预警研究的一种新思路。

The financial early warning establishes the forecasting model to predict the risk by analyzing the relevant indicators of the enterprise,and provides the basis for the related decision-making of the stakeholders,so research on the efficiency of early warning becomes a critical point. Taking the relevant data of 90 manufacturing enterprises to build a probabilistic neural network model for early warning research,the particle swarm optimization algorithm is introduced to get a promotion of the predict efficiency. According to the empirical analysis,the prediction accuracy of the pre-optimization model without particle swarm optimization is 87.5%,while the optimized one is 93.75%. It is shown that the particle swarm optimization algorithm is feasible for the optimization of neural networks,which can provide a new way for the financial early warning research of listed companies.