PSO算法优化BP神经网络的金融风险预警研究

金融危机是一个非线性的复杂过程，BP神经网络对非线性系统具有很强的模拟能力。针对BP神经网络有收敛速度慢、易陷入局部极小值和振荡等缺点，利用改进的PSO算法优化BP神经网络的权值和阈值，能有效地改善BP神经网络的缺点。对金融风险实例分析的结果表明，综合改进的BP算法相对于BP神经网络算法能明显加快网络的收敛时间，具有较快的收敛速度和较高的诊断准确度，用于金融风险预警是可行的，证实了该方法具有一定的实际应用价值。

Study on financial risk warning based on BP neural network optimized by the improved PSO

The financial crisis is a complex process of nonlinear, the BPcapability of nonlinear simulation system. Because BP neural network has slowfallingneural network has strongconvergence speed, easilyinto local minima and oscillations cannot be applied. Through using the optimized BP neuralnetwork by PSO algorithm to improve the weight and threshold value, can effectively improve theshortcoming of BP neural network. The results of the financial risk analysis show that, the BP algorithmcombines the improved BP neural network algorithm can accelerate the convergence time relative to thediagnosis, which has faster convergence speed and higher accuracy, for the financial risk early warning isfeasible, have proved that this method has a certain practical application value.