基于复杂网络的证券市场智能建模与分析  

Intelligent Modeling and Analysis of Stock Market Based on Complex Network

作　　者：李双宏 舒子宸 肖雅雯 LI Shuanghong;SHU Zichen;XIAO Yawen(Department of Research and Development,Orient Securities Company Limited,Shanghai 200010;Melman School of Public Health,Columbia University,New York 10032)

机构地区：[1]东方证券股份有限公司系统研发总部,上海200010 [2]哥伦比亚大学梅尔曼公共卫生学院,纽约10032

出　　处：《计算机与数字工程》2021年第12期2414-2419,共6页Computer & Digital Engineering

摘　　要：股票市场作为金融系统的重要组成部分,是一个典型的具有结构复杂性和节点复杂性的复杂网络系统。作为拆分和了解复杂网络的有力工具,社团结构分析被广泛应用于社交网络、物流网络等多种复杂网络系统,并取得了突破性成果。论文采用Pearson相关系数来度量中国A股市场中股票价格波动的相关关系,构建股票市场加权网络,利用改进型社团相似性指标,选定了股票市场时序动态加权网络的步长与社团划分算法,并对社团结构进行了简要分析。Stock market is an indispensable part of the financial system,and a typical complex network system with apparently structural complexity and vertex complexity. As a powerful tool for splitting and understanding complex network,community structure analysis is introduced into social networks,logistics networks and so on,which brought relevant research frontiers a lot of breakthroughs. This article measures the correlation between the price fluctuation of stocks by Pearson correlation coefficient,managing to construct an authorized network about the whole stock market since unauthorized network of stock market contains many isolated nodes which are not available for further research. Then the step length of the authorized network of Chinese stock market are determined by introducing the enhanced community similarity index as the criteria,while found the most appropriate communities partition algorithm from the perspective of modeling and analysis requirement. Comparing to relevant research,this article upgrade the scale of network to the whole stock market for the first time,and bring up an new approach for parameters tuning during the process of network constructing.

关 键 词：股票市场 复杂网络 社团分析 

分 类 号：F224[经济管理—国民经济]