Study of Financial Warning Ensemble Model for Listed Companies Based on Unbalanced Classification Perspective

Study of Financial Warning Ensemble Model for Listed Companies Based on Unbalanced Classification Perspective

Wei Cong (School of Management Science and Engineering, Nanjing University of Information Science and Technology, Nan Jing, China)
Copyright: © 2020 |Pages: 17
DOI: 10.4018/IJIIT.2020010103

Abstract

Using the ensemble learning method to mine valuable information from a sea of financial data accumulated on the market of financial securities is very important for studying data processing. On the basis of financial data from A-share companies listed on Shanghai Stock Market, this article takes the perspective of unbalanced classification of ST stocks to carry out a study of the construction of a financial warning model for the listed companies. In our experiment, HDRF (HDRandom Forest, Hellinger Distance based Random Forest), ensemble classification models of Bagging, AdaBoost, and Rotation Forest, which take Hellinger distance decision tree (HDDT) as the base classifier, and the ensemble classification model which takes the C4.5 decision tree as the base classifier, are compared in respect of both the area under the ROC curve and the F-measure. As shown in the experimental results, the HDRF and the HDDT based classifier, as an ensemble method, are effective for financial data of listed companies.
Article Preview
Top

1. Introduction

A key direction for studying the integration of analytical processing and ensemble learning of financial data is to accurately mine valuable information from a sea of real-time, incomplete and unbalanced data on China’s securities market. The objective of this article is to introduce different ensemble learning methods from the perspective of unbalanced classification to explore the ensemble classification model suitable for the corresponding data structure. There has been no research in using the after mentioned algorithm for financial warning prediction. The contribution of this paper is to build a relatively effective classification model to improve the forecasting ability of ST shares in financial data of listed companies, which has certain theoretical and practical significance for the construction and application of financial warning model under background of financial big data.

1.1. Theories Relating to Financial Warning for Listed Companies

Financial warning, as an important content of the management of a company’s financial risks, for a listed company, the financial warning means to make statistics of the company’s data such as financial statements and operating plans to predict its operational and financial activities as well as to find out operational and financial risks in its operation and management activities, so as to inform shareholders and managers with warning information.

A number of scholars have already made systematic studies on financial warning. Among them, Qin (2012) applied a Logistic regression analysis model to empirically study the selection of financial warning variables for companies listed on Shanghai and Shenzhen A-share markets whose stocks were, for the first time, prefixed with “ST”. Xiao and Yang (2018) constructed L1/2-regularized Logistic Regression model and took ST and non-ST manufacturing companies in Shanghai and Shenzhen stock markets as the research objects for comparative prediction. Song et al. (2019) used Cox regression to build financial crisis warning model for minor enterprises to achieve relatively accurate prediction of healthy enterprises. By using the annual reporting data of two sets of 37 pairs of “A” shares, Wang (2008) built a warning model for listed companies based on a multiple regression model and checked the accuracy as well, and he also gave his opinion and suggestions about the financial warning system. Li et al. (2013) adopt multiple discriminant analysis to modify a Z-Score baseline model, and establish a financial distress early warning model applicable to listed real estate companies in China. The findings indicate that the average accuracy of the financial distress early warning model reaches higher than 90%. Wang et al. (2018) combined PSO-BP and FOA-BP neural network model to select data of high-tech enterprises to effectively predict the financial crisis of enterprises. As indicated in the above studies, in most of the empirical studies about the financial warning for listed companies, a reasonable statistical analysis model is usually built based on financial data to check the validity and make a systematic analysis of the warning model. This paper has studied financial warning ensemble model for listed Companies based on Unbalanced Classification Perspective.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 16: 4 Issues (2020): 1 Released, 3 Forthcoming
Volume 15: 4 Issues (2019)
Volume 14: 4 Issues (2018)
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing