论文部分内容阅读
网络分析方法为系统分析股票市场的复杂性提供了有效方法。本文利用最小生成树和分层树聚类分析法的思想,引入符号时间序列分析法(STSA),通过构建股票市场的拓扑结构来分析经济意义上的聚类特征。首先,将收益序列转化为符号序列,并在此基础上计算出符号序列的欧氏距离;其次,利用Kruskal的最小生成树算法,构建出证券组合的亚超度量距离矩阵,并把亚超度量距离矩阵映射为分层树结构;最后,对上证50指数成分股进行了实证分析。
Network analysis methods provide an effective way to systematically analyze the complexity of the stock market. In this paper, using the idea of minimum spanning tree and hierarchical tree clustering analysis, the introduction of symbolic time series analysis (STSA), by analyzing the topological structure of the stock market to analyze the economic characteristics of the cluster. First, the income sequence is converted into a symbol sequence, and based on this, the Euclidean distance of the symbol sequence is calculated. Secondly, the Kruskal minimum spanning tree algorithm is used to construct the sub-metric distance matrix of the security portfolio, The distance matrix is mapped into a hierarchical tree structure. Finally, empirical analysis is conducted on the constituent stocks of the SSE 50 Index.