基于分割提速法的并行股票预测研究与仿真

摘要

关于股票准确预测问题,针对股票预测中计算时间复杂度大,数据库操作速度慢等缺点,为提高数据挖掘的速度和效率,提出了一种分割提速法的并行股票预测模型.模型特点采用广播拓扑结构,使用多线程并行计算方法,将计算量平均分配给所有参与计算的计算机,同时又使用网络编程技术实时同步回收结果,从而有效地缩短了股票关联规则的计算时间.实验结果表明,上述方法有效地减少了算法时间复杂度,较大程度地提高了股票预测的效率,从而为股票投资者提供有力的帮助.%To overcome the shortcomings such as large computation complexity in stock predication, slow data-base operation speed etc. and to improve the speed and efficiency of data mining, a parallel stock prediction model based on apriori partition speed method was proposed. It made good use of broadcasting topology architecture and multi-thread parallel computing, which can well-distribute the computing on multi-computers. Meanwhile, it gave real-time result collecting by using socket programming, thus reduced the computing time effectively. The experimen-tal results show that the method reduces the time complexity of algorithm effectively, improves greatly the efficiency of stock predication, and offers a great help to the stock trader.