Abstract: As the pillar industry of the financial industry, securities have an important impact on the capital market. As the main department of securities issuance and trading, securities companies play an important role in China’s capital market construction, economic transformation, system reform and other aspects. The business structure of securities companies, that is, the various businesses operated by securities companies and the combination relationship between them, determines the profit level of securities companies, and is also one of the important factors affecting the risk level of securities companies. Therefore, the selection of business structure of securities companies is a very important theoretical and practical problem for the survival and development of securities companies, and it is also one of the hot issues in the current securities industry.
Through analysis and research, we find that the integration problem in the selection of business structure of securities companies is actually a combination problem of the business of securities companies according to some requirements, which is different from the securities portfolio problem, and just coincides with the relevant content of cluster analysis application research. Therefore, this paper considers the clustering method in clustering analysis, and quantitatively studies the integration problem in the selection of business structure of securities companies.
Based on the qualitative research results of the business structure selection and integration of securities companies, the practical quantitative integration method is given. For the securities business structure with multi-business as the core, the interactive self-organization data algorithm is used to soft-divide the securities business collection, so as to realize the problem of business structure selection and integration of securities companies that cannot be specifically solved in the qualitative research, combined with the application calculation examples.
In short, this paper studies the integration model and algorithm of the business structure selection of securities companies. For the securities business structure with multi-business as the core, the interactive self-organization data algorithm is used to soft-divide the securities business collection. In this way, the integration problem of business structure selection of securities companies that cannot be specifically solved in qualitative research is realized, and the transformation from qualitative research to quantitative research is completed.

Key words: securities companies, integration of securities business structure, interactive self-organizing data algorithm

摘要： 借鉴证券公司业务结构选择整合问题的有关定性研究成果,给出实用的量化整合方法。对多业务为核心的证券业务结构,使用交互式自组织数据算法对证券业务集合进行软划分,实现定性研究中不能具体解决的证券公司业务结构选择整合问题。此外,结合H证券公司的应用算例,将软划分法与经典聚类划分法进行对比分析,以此说明本文所用方法的合理性。

关键词: 证券公司, 证券业务结构整合, 交互式自组织数据算法