人工神经网络在证券价格预测中的应用

证券市场中成功的交易模式是可以模仿及学习的.证券价格走势实质是一种复杂时序函数.人工神经网络是在模仿人脑处理问题过程中发展起来的新型智能信息处理系统,人工神经网络可以通过调节连接权值以任意精度逼近任何连续函数,因此也可以逼近证券价格随时间变换这种函数.文中采用基于BP模型的神经网络,用BP算法和遗传算法来训练网络权值,同时也采用了动量法和学习率自适应调整相结合的策略,对证券市场的价格进行建模和预测,结果表明,此模型具有较好的学习、泛化能力,对股票市场或其他类似的非线性经济系统的走势预测决策具有较好的效果.

Application of Artificial Neural Networks in Price Forecast of Securities Market

The successful trading mode in securities market can be learned and copied.The price curve of securities is a complex time-series function.Artificial Neural Network(ANN) is a new intelligent information processing theory developed in the process of imitating brain to solve problem.Artificial neural network can ximate any continuous function precisely by adjusting connection weights,so it can also approximate the stock price function which changes along with time.This paper presents a method for securities market modeling and forecasting via neural network based on BP model,in which the genetic algorithm is used to train the connection weigths of the neural network,and the way of momentum and self-adaptive learning rate are also used.It has shown by the modeling and forecasting results about securities market price that the method has learning properties and reinforcement.This method is useful for modeling and forecasting securities market or some other similar nonlinear economic system.