基于HPSO-BP神经网络的个人信用评估

Personal Credit Evaluation based on HPSO-BP Neural Network

为了解决BP神经网络和标准PSO-BP神经网络模型收敛慢、易陷入局部最优值等问题,引入改进的粒子群算法HPSO,提出了基于HPSO-BP神经网络的信用评估模型。在PyCharm环境下,利用德国个人信用数据集,分别比较了BP神经网络模型、标准PSO-BP神经网络模型和文中的HPSO-BP神经网络模型。实验结果表明,基于HPSO-BP神经网络的评估模型在收敛速度和准确度上都优于另外两个模型。

To solve the problems such as slow convergence,local optimal of BP neural network and standard PSO-BP neural network credit evaluation models,this paper introduces an improved particle swarm algorithm HPSO,and propo-ses a credit evaluation model based on HPSO-BP neural network.Under PyCharm,by using the German personal credit data set,a comparing experiment is done on the BP neural network,the standard PSO-BP neural network and the HPSO-BP neural network.The results show that the credit evaluation model based on HPSO-BP neural network has better con-vergence speed and higher accuracy than the other two models.