遗传算法和神经网络耦合的金融预测系统  

Financial Prediction System Based on Combinaton of Genetic Algorithm with BP Neural Network

作　　者：宋晓勇[1] 陈年生[1] 

机构地区：[1]上海电机学院电子信息学院,上海201306

出　　处：《上海交通大学学报》2016年第2期313-316,共4页Journal of Shanghai Jiaotong University

基　　金：上海电机学院重点学科项目(13XKJ01)资助

摘　　要：神经网络具有强大处理非线性系统的能力和映射能力,在财务预警和金融预测中得到广泛应用.神经网络与遗传算法耦合的金融参数预测系统(GA-BP系统)是利用智能模拟算法,算法要点是遗传算法对神经网络预测金融系统拓扑结构层间权系值进行优胜劣汰演化,本文证明二者耦合能提高网络系统的效率和预测精度,实现了两种智能模拟方法的集成耦合.Neural network can deal with nonlinear systems effectively and is applied widely on financial early warning and financial forecast. Neural network coupled with genetic algorithms is an intelligent simulaton algorithm in predicting financial parameter system. Genetic algorithm predicts the coefficient of the topology layers in neural network predicting the financial parameter system. This paper predicts business debt ratio. It proves that the combination of genetic algorithm with BP neural network can improve the efficiency and the predicting precision of network system.

关 键 词：神经网络 遗传算法 金融预测 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]