基于BP神经网络的中国上市公司现金股利识别模型构建——以农业上市公司为例

Establishment of the Cash Dividend Identification Model for Chinese Listed Companies Based on the BP Neural Network——Taking Listed Agricultural Companies as an Example

双重委托代理关系加上激励、约束有效机制的缺乏构成中国上市公司股权结构的特殊性,这种特殊的股权结构使得中国上市公司现金股利分配一直以来很不规范,严重地侵害了广大中小投资者的利益。本研究将BP神经网络技术运用于上市公司现金股利,探讨通过建立BP现金股利识别模型实现对上市公司现金股利分配的识别预测,解决目前广大中小投资者对上市公司现金股利的判断主要集中在事后,缺少事前识别和预测的问题。本研究基于中国农业上市公司2011-2017的数据,围绕股权结构选取六项指标,构建了BP现金股利识别模型,分析了利用模型识别现金股利的基本原理,并对模型的应用效果进行了识别检验。本文为研究和开展上市公司现金股利识别跟踪提供了一种可行路径,也为人工智能与现金股利结合探讨提供了一种思路和范例。

The double principal-agent relationships plus the lack of efficient incentive and restraint mechanisms constitute the particularity of the equity structure of listed companies in China.This special equity structure makes the cash dividend distribution of listed companies very irregular all the time,which seriously infringes the interests of the majority of small and medium-sized investors.In this study,the BP neural network technology is applied to cash dividends of listed companies.By establishing a BP cash dividend identification model,it is feasible to identify and predict cash dividends distribution of listed companies,so as to solve the problem that the majority of small and medium-sized investors mainly judge cash dividends of listed companies in the post operation and lack prior identification and prediction.Based on the data of China's agricultural listed companies from 2011 to 2017,six indicators around the ownership structure are selected,a BP cash dividend identification model is constructed,the basic principle of using the model to identify cash dividends is analyzed,and the application effect of the model is tested.Accordingly,a feasible path for the research and development of cash dividend identification and tracking of listed companies as well as a novel thought and example is provided herein for the combination of artificial intelligence and cash dividend.