【摘要】 本文研究了基于神经网络的股票预测方法．针对目前存在的问题，提出了联合Davidon最小二乘算法及遗传算法来综合训练网络结构和权值的新方法．经对上证指数的模拟预测表明，通过合理地选取参数，可获得满意的预测效果．更多还原

【Abstract】 This paper is focused on using neural networks (NN) method to predict stock market- Firstly, the complexity of stock market and the existed prediction methods are summarized. Secondly, a noval method for optimizing the strucutre and weights of NN is proposed, which combines genetic algorithm (GA) and Davidon,s least squares algorithm. Finally, Shanghai stock indexes are fitted and predicted. The good results confirms the performance of the method.更多还原