摘要
为了对中国股票进行数据分析,针对股票市场中的股票价格之间的相关性,提出了一种基于Apriori算法的改进算法。算法在垂直数据表示方式上执行广度优先搜索和交叉计数,充分利用了垂直数据表示和交叉计数的高效优势,以及Apriori算法的剪枝策略,减少了计数的候选项集的数量。并对两种算法进行了性能比较,改进的Apriori算法的运行速度较Apriori算法有明显的提高。最后将新算法应用于股票分析仿真系统,仿真结果表明,改进算法能够得到有意义的规则,快速发现股票之间的涨跌关系,为投资者提供了实时、准确的股票买入还是卖出的决策支持。
In order to analysis China stock market, amed at the relativity among some stock prices, an adapted association rule mining algorithm is given based on Apriori algorithm. The new algorithm is implemented with vertical data layout, breadth first searching, and intersecting. It makes use of the efficiency of vertical data layout and intersecting, and prunes candidate frequent item sets like Apriori. The new algorithm is experimentally compared against Apriori, and it is found that significant progress has been made in runtime on our test database. Finally,the new algorithm is applied in simulation of stock analysis system. The result shows that the new algorithm is able to mine reasonable rules, find out the relationship among stocks, and provide timely, accurate decision support for investors.
出处
《计算机仿真》
CSCD
北大核心
2010年第6期334-337,共4页
Computer Simulation
关键词
关联规则
股票分析
交叉计数
算法
Association rule
Stock analysis
Intersecting
Algorithm
