基于优化BP算法的上证指数预测分析

上证指数变化是股市投资者重点关注的指标之一。文章基于BP神经网络来建立预测模型,结合遗传算法优化BP网络存在的不足,然后通过组合不同的辅助技术指标优化训练样本,探寻对上证指数预测准确度较高的混合算法。实验结果表明,优化后的混合算法能较好的预测上证指数未来一段时间的走势变化,为投资者提供决策辅助。

Prediction analysis of SSE composite index based on optimized BP algorithm

SSE （Shanghai Stock Exchange） composite index change is one of the indicators that the stock market investors fo- cus on. Based on the BP neural network, a prediction model was constructed, and then BP network was optimized via the genetic algorithm. Afterwards, the training samples were optimized through the combination of different auxiliary technical indicators to seek for a hybrid algorithm that could forecast SSE composite index more accurately. The experimental results show that the optimized hybrid algorithm can better predict the change trend of SSE index in a future period of time, which provides investors with a decision aiding.