基于量子遗传算法的粗糙集属性约简的研究

摘要

This paper studies the stock trend forecast by using the rough set theory.Besides applying existing reduction methods,it also uses a new reduction method based on quantum computing and genetic algorithm for attribute reduction.Simultaneously,a new quantum selection revolving door adjusting tactics which can improve the speed of convergence has been used.The tests prove that the new method has significantly improved in speed and astringency the stock trend forecast model,which is of great significance to some extent.%将粗糙集理论应用于股市的分析与研究,除了采用已有的约简方法进行试验之外,还引入了量子计算与遗传算法相结合的方法来进行粗糙集的属性约简.与其他约简算法不同的是,该算法采用量子旋转门策略来达到全局最优搜索和较高的收敛速度.最后,通过具有代表性的股票数据证明了该方法的有效性和高效性.