基于关联规则挖掘算法的A股股票联动性分析

Association rule mining algorithm based analysis of A-share stock linkage

为了更好地观察国内A股间的联动性,针对股票效应的滞后性问题,提出了一种基于时序的改进关联规则挖掘算法Gap-Apriori.实验采用Apriori、FP-growth、Eclat、Gap-Apriori 4种关联规则挖掘算法,对我国2007年到2021年间的A股交易数据进行了关联分析.实验结果表明,Apriori算法较其他3种算法更适用于高维股票数据挖掘,改进算法Gap-Apriori能够分析任意周期内的股票联动状态,有效地提高了算法的运行效率.

As one of the hotspots in the financial field,stock analysis has attracted a large number of researchers to join it.Finding patterns in massive stock historical data and predicting stock trends has become a hot research topic.In order to better observe the linkage between domestic A shares,aiming at the lag of stock comovement,an improved association rule mining algorithm gap Apriori based on time series is proposed.The experiment uses four association rule mining algorithms,namely,Apriori,FP-growth,Eclat,and Gap-Apriori,to conduct an association analysis on the A-share transaction data of China from 2007 to 2021.The experimental results show that the Apriori algorithm is more suitable for high-dimensional stock data mining than the other three algorithms.The improved algorithm Gap-Apriori can analyze the stock linkage status in any period,which effectively improves the efficiency of the algorithm.