摘要
本文提出一种GA-BP算法对金融股指进行预测研究,该算法结合遗传算法具有全局寻优的优点,能够克服BP算法进行预测时易陷入局部极小值和收敛速度慢的缺点;以股票市场的上证指数为例,通过与BP算法和LM-BP算法预测效果进行比较分析,证实了本文提出的算法在对上证指数进行预测时更为有效。
A GA-BP algorithm (Genetic Algorithm-Back Propagation Algorithm) is formed in prediction of financial stock index. According to the advantage of the globe optimal searching of genetic algorithm, this algorithm can overcome problems of local minima and low convergence speed in BP algorithm. Comparing with BP algorithm and LM-BP algorithm, this algorithm is effective in prediction of SSE Composite Index.
出处
《技术经济与管理研究》
北大核心
2009年第6期16-18,43,共4页
Journal of Technical Economics & Management
关键词
股票指数预测
神经网络
BP算法
GA算法
stock index prediction
neural network
BP algorithm
genetic algorithm