Application of Stacking-Bagging-Vote multi-source information fusion model for financial early warning

doi:10.11772/j.issn.1001-9081.2021020306

Journal of Computer Applications ›› 2022, Vol. 42 ›› Issue (1): 280-286.DOI: 10.11772/j.issn.1001-9081.2021020306

• Frontier and comprehensive applications • Previous Articles Next Articles

Application of Stacking-Bagging-Vote multi-source information fusion model for financial early warning

Lu ZHANG, Jiapeng LIU(), Dongmei TIAN

College of Economics and Management，China Jiliang University，Hangzhou Zhejiang 310018，China

Received:2021-03-02 Revised:2021-06-21 Accepted:2021-06-23 Online:2022-01-11 Published:2022-01-10
Contact: Jiapeng LIU
About author:ZHANG Lu， born in 1995， M. S. candidate. Her research interests include corporate finance， data mining.
LIU Jiapeng， born in 1969， Ph. D.， professor. His research interests include financial information.
TIAN Dongmei， born in 1996， M. S. candidate. Her research interests include intelligent investment， quantitative trading.
Supported by:
National Social Science Foundation of China(18BGL224)

基于Stacking-Bagging-Vote多源信息融合模型的财务预警应用

张露, 刘家鹏(), 田冬梅

中国计量大学经济与管理学院，杭州 310018

通讯作者: 刘家鹏
作者简介:张露（1995—），女，浙江宁波人，硕士研究生，主要研究方向：公司金融、数据挖掘
刘家鹏（1969—），男，山东日照人，教授，博士，CCF会员，主要研究方向：金融信息；田东梅（1996—），女，河南南阳人，硕士研究生，主要研究方向：智能投资、量化交易。
基金资助:
国家社会科学基金资助项目(18BGL224)

Abstract

Abstract:

Ensemble resampling technology can solve the problem of imbalanced samples in financial early warning research to some extent. Different ensemble models and different ensemble resampling technologies have different suitabilities. It is found in the study that Up-Down ensemble sampling and Tomek-Smote ensemble sampling were respectively suitable for Bagging-Vote ensemble model and Stacking fusion model. Based on the above， a Stacking-Bagging-Vote （SBV） multi-source information fusion model was built. Firstly， the Bagging-Vote model based on Up-Down ensemble sampling and the Stacking model based on Tomek-Smote sampling were fused. Then， the stock trading data were added and processed by Kalman filtering， so that the interactive fusion optimization of data level and model level was realized， and the SBV multi-source information fusion model was finally obtained. This fusion model not only has a great improvement in the prediction performance by taking into account prediction accuracy and prediction precision simultaneously， but also can select the corresponding SBV multi-source information fusion model to perform the financial early warning to meet the actual needs of different stakeholders by adjusting the parameters of the model.

Key words: financial early warning, multi-source information fusion, ensemble resampling technology, Stacking-Bagging-Vote (SVB) model, Kalman filtering

摘要：

集成重采样技术可以在一定程度上解决财务预警研究中样本的不平衡性难题，而不同的集成模型与不同的重采样集成技术有不同的适配性。研究发现，Up-Down集成采样与Tomek-Smote集成采样分别适配于Bagging-Vote集成模型和Stacking融合模型。基于此，构建了Stacking-Bagging-Vote （SBV）多源信息融合模型。首先，将基于Up-Down集成采样的Bagging-Vote模型与基于Tomek-Smote采样的Stacking模型进行融合；然后，加入股票的交易数据，并对该数据用卡尔曼滤波进行处理，从而形成数据层次和模型层次的交互式融合优化；最终，得到SBV多源信息融合模型。该融合模型不仅在预测性能上有了较大的提升，能较好地兼顾模型的预测准确度和预测精确率，并且可以根据利益相关者的实际需要，通过调整模型参数，来选择对应的SBV多源信息融合模型进行财务预警预测。

关键词: 财务预警, 多源信息融合, 集成重采样技术, Stacking-Bagging-Vote模型, 卡尔曼滤波

CLC Number:

TP181

Lu ZHANG, Jiapeng LIU, Dongmei TIAN. Application of Stacking-Bagging-Vote multi-source information fusion model for financial early warning[J]. Journal of Computer Applications, 2022, 42(1): 280-286.

张露, 刘家鹏, 田冬梅. 基于Stacking-Bagging-Vote多源信息融合模型的财务预警应用[J]. 《计算机应用》唯一官方网站, 2022, 42(1): 280-286.

Figures/Tables 10

References 19

1	孟小峰，杜治娟. 大数据融合研究：问题与挑战［J］. 计算机研究与发展， 2016， 53（2）：231-246. 10.7544/issn1000-1239.2016.20150874
	MENG X F， DU Z J. Research on the big data fusion： issues and challenges［J］. Journal of Computer Research and Development， 2016， 53（2）：231-246. 10.7544/issn1000-1239.2016.20150874
2	BARBOZA F， KIMURA H， ALTMAN E. Machine learning models and bankruptcy prediction［J］. Expert Systems with Applications， 2017， 83： 405-417. 10.1016/j.eswa.2017.04.006
3	杨剑锋，乔佩蕊，李永梅，等. 机器学习分类问题及算法研究综述［J］. 统计与决策， 2019， 35（6）：36-40. 10.13546/j.cnki.tjyjc.2019.06.008
	YANG J F， QIAO P R， LI Y M， et al. A review of machine-learning classification problems and algorithms［J］. Statistics and Decision， 2019， 35（6）：36-40. 10.13546/j.cnki.tjyjc.2019.06.008
4	HE H B， GARCIA E A. Learning from imbalanced data［J］. IEEE Transactions on Knowledge and Data Engineering， 2009， 21（9）：1263-1284. 10.1109/tkde.2008.239
5	CATENI S， COLLA V， VANNUCCI M. A method for resampling imbalanced datasets in binary classification tasks for real-world problems［J］. Neurocomputing， 2014， 135：32-41. 10.1016/j.neucom.2013.05.059
6	VEGANZONES D， SÉVERIN E. An investigation of bankruptcy prediction in imbalanced datasets［J］ Decision Support Systems， 2018， 112：111-124. 10.1016/j.dss.2018.06.011
7	TAHIR M A， KITTLER J， MIKOLAJCZYK K， et al. A multiple expert approach to the class imbalance problem using inverse random under sampling［C］// Proceedings of the 2009 Multiple Classifier Systems， LNCS5519. Berlin： Springer， 2009：82-91.
8	CHAWLA N V， BOWYER K W， HALL L O， et al. SMOTE： synthetic minority over-sampling technique［J］. Journal of Artificial Intelligence Research， 2002， 16：321-357. 10.1613/jair.953
9	QIAN Y， LIANG Y C， LI M， et al. A resampling ensemble algorithm for classification of imbalance problems［J］. Neurocomputing， 2014， 143：57-67. 10.1016/j.neucom.2014.06.021
10	ZOU H， HASTIE T. Regularization and variable selection via the elastic net［J］. Journal of the Royal Statistical Society： Series B （Statistical Methodology）， 2005， 67（2）：301-320. 10.1111/j.1467-9868.2005.00503.x
11	BREIMAN L. Random forests［J］. Machine Learning， 2001， 45（1）：5-32. 10.1023/a:1010933404324
12	CHEN T Q， GUESTRIN C. XGBoost： a scalable tree boosting system［C］// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2016：785-794. 10.1145/2939672.2939785
13	徐继伟，杨云. 集成学习方法：研究综述［J］. 云南大学学报（自然科学版）， 2018， 40（6）：1082-1092. 10.1007/978-3-319-32903-1_103-1
	XU J W， YANG Y. A survey of ensemble learning approaches［J］. Journal of Yunnan University （Natural Sciences Edition）， 2018， 40（6）：1082-1092. 10.1007/978-3-319-32903-1_103-1
14	WOLPERT D H. Stacked generalization［J］. Neural Networks， 1992， 5（2）：241-259. 10.1016/s0893-6080(05)80023-1
15	王奕森，夏树涛. 集成学习之随机森林算法综述［J］. 信息通信技术， 2018， 12（1）：49-55. 10.3969/j.issn.1674-1285.2018.01.009
	WANG Y S， XIA S T. A survey of random forests algorithms［J］. Information and Communication Technologies， 2018， 12（1）：49-55. 10.3969/j.issn.1674-1285.2018.01.009
16	连克强. 基于Boosting的集成树算法研究与分析［D］. 北京：中国地质大学（北京）， 2018：12-32. 10.30919/es8d689
	LIAN K Q. The study and application of ensemble of trees based on Boosting［D］. Beijing： China University of Geosciences， Beijing， 2018：12-32. 10.30919/es8d689
17	朴杨鹤然，任俊玲. 基于Stacking的恶意网页集成检测方法［J］. 计算机应用， 2019， 39（4）：1081-1088. 10.11772/j.issn.1001-9081.2018091926
	PIAOYANG H R， REN J L. Malicious webpage integrated detection method based on Stacking ensemble algorithm［J］. Journal of Computer Applications， 2019， 39（4）： 1081-1088. 10.11772/j.issn.1001-9081.2018091926
18	丁岚，骆品亮. 基于Stacking集成策略的P2P网贷违约风险预警研究［J］. 投资研究， 2017， 36（4）：41-54.
	DING L， LUO P L. Research on default risk early-warning in P2P lending based on Stacking ensemble strategy［J］. Review of Investment Studies， 2017， 36（4）：41-54.
19	任慧玉，陈景华，任凌玉，等. 卡尔曼滤波方法在β估计中的应用［J］. 河南大学学报（自然科学版）， 2004， 34（2）：10-14. 10.3969/j.issn.1003-4978.2004.02.002
	REN H Y， CHEN J H， REN L Y， et al. The application of the Kalman filtering in the estimation of the asset’s systemic risk［J］. Journal of Henan University （Natural Science）， 2004， 34（2）：10-14. 10.3969/j.issn.1003-4978.2004.02.002

预测类别	真实类别
预测类别	P	N
P	TP	FN
N	FP	TN

预测类别	真实类别
预测类别	P	N
P	TP	FN
N	FP	TN

类型	样本数	占比/%	Y取值
合计	2 936	100.00
AB	31	1.05	1
AD	51	1.74	1
AX	2	0.07	1
AA	2 852	97.14	0

类型	样本数	占比/%	Y取值
合计	2 936	100.00
AB	31	1.05	1
AD	51	1.74	1
AX	2	0.07	1
AA	2 852	97.14	0

采样算法	集成模型	召回率	精确率	F1值	G-mean
集成Up-Down	BV-EN	0.761 9	0.113 3	0.197 2	0.793 1
	BV-RF	0.785 7	0.370 8	0.503 8	0.873 2
	BV-XGBoost	0.845 2	0.282 9	0.423 9	0.891 1
	BV-Models	0.869 0	0.298 0	0.443 8	0.904 4
	Stacking-DT	0.833 3	0.123 0	0.214 4	0.829 2
	Stacking-SVM	0.940 5	0.092 7	0.168 8	0.834 7
	Stacking-LR	0.881 0	0.129 6	0.226 0	0.853 3
SMOTE	BV-EN	0.535 7	0.211 3	0.303 0	0.710 0
	BV-RF	0.821 4	0.249 1	0.382 3	0.872 7
	BV-XGBoost	0.821 4	0.227 7	0.356 6	0.868 4
	BV-Models	0.654 8	0.348 1	0.794 4	0.454 5
	Stacking-DT	0.904 8	0.160 3	0.272 4	0.882 3
	Stacking-SVM	0.928 6	0.139 3	0.242 2	0.879 8
	Stacking-LR	0.904 8	0.191 9	0.316 7	0.896 2
Tomek-Smote	BV-EN	0.571 4	0.151 4	0.239 4	0.738 5
	BV-RF	0.845 2	0.229 8	0.361 3	0.880 2
	BV-XGBoost	0.809 5	0.191 0	0.309 1	0.853 1
	BV-Models	0.607 1	0.377 8	0.767 6	0.465 8
	Stacking-DT	0.881 0	0.183 2	0.303 3	0.882 6
	Stacking-SVM	0.916 7	0.138 5	0.240 6	0.874 4
	Stacking-LR	0.916 7	0.201 0	0.329 8	0.904 6

Application of Stacking-Bagging-Vote multi-source information fusion model for financial early warning

基于Stacking-Bagging-Vote多源信息融合模型的财务预警应用

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 10

References 19

Related Articles 15

Recommended Articles

Metrics

排序指标	融合模型	召回率	精确率	F1值	G-mean
召回率	SBV-MF-R1	0.976 2	0.132 7	0.233 6	0.894 1
	SBV-MF-R2	0.976 2	0.119 7	0.213 3	0.882 4
	SBV-M-R1	0.964 3	0.143 9	0.250 4	0.897 6
	SBV-M-R2	0.964 3	0.142 9	0.248 8	0.896 9
	SBV-S-R1	0.964 3	0.130 6	0.230 1	0.887 6
	SBV-S-R2	0.964 3	0.120 5	0.214 3	0.878 5
精确率	SBV-S-P1	0.916 7	0.269 2	0.416 2	0.921 7
	SBV-S-P2	0.857 1	0.268 7	0.409 1	0.894 2
	SBV-M-P1	0.904 8	0.255 9	0.399 0	0.913 6
	SBV-M-P2	0.928 6	0.242 2	0.384 2	0.921 5
	SBV-MF-P1	0.940 5	0.161 2	0.275 3	0.898 2
	SBV-MF-P2	0.964 3	0.141 9	0.247 3	0.896 2
G-mean	SBV-S-G1	0.928 6	0.246 8	0.390 0	0.922 6
	SBV-S-G2	0.916 7	0.269 2	0.416 2	0.921 7
	SBV-M-G1	0.928 6	0.242 2	0.384 2	0.921 5
	SBV-M-G2	0.940 5	0.212 4	0.346 5	0.918 9
	SBV-MF-G1	0.940 5	0.161 2	0.275 3	0.898 2
	SBV-MF-G2	0.964 3	0.141 9	0.247 3	0.896 2

[1]	Shoulong JIAO, Youxiang DUAN, Qifeng SUN, Zihao ZHUANG, Chenhao SUN. Knowledge representation learning method incorporating entity description information and neighbor node features [J]. Journal of Computer Applications, 2022, 42(4): 1050-1056.
[2]	Xingshuo DING, Xiang LI, Qian XIE. Enterprise portrait construction method based on label layering and deepening modeling [J]. Journal of Computer Applications, 2022, 42(4): 1170-1177.
[3]	SHANG Jiandong, LI Panle, LIU Runjie, LI Runchuan. Recommendation model of taxi passenger-finding locations based on weighted non-homogeneous Poisson model [J]. Journal of Computer Applications, 2018, 38(4): 923-927.
[4]	ZHANG Hengde, XIAN Yunhao, XIE Yonghua, YANG Le, ZHANG Tianhang. Haze forecast based on time series analysis and Kalman filtering [J]. Journal of Computer Applications, 2017, 37(11): 3311-3316.
[5]	PAN Lei, ZHOU Huan, WANG Minghui. Real-time detection method of abnormal event in crowds [J]. Journal of Computer Applications, 2016, 36(6): 1719-1723.
[6]	SHI Gang, ZHAO Wei, LIU Shanshan. Battery SOC estimation based on unscented Kalman filtering [J]. Journal of Computer Applications, 2016, 36(12): 3492-3498.
[7]	REN Hongge, XIANG Yingfan, LI Fujin. Autonomous developmental algorithm for intelligent robot based on intrinsic motivation [J]. Journal of Computer Applications, 2015, 35(9): 2602-2605.
[8]	QU Haicheng, SHAN Xiaochen, MENG Yu, LIU Wanjun. Improved TLD target tracking algorithm based on automatic adjustment of surveyed areas [J]. Journal of Computer Applications, 2015, 35(10): 2985-2989.
[9]	ZHU Wenchao XU Dezhang. Model error restoration for lower E-type membrane of six-axis force sensor based on adaptive Kalman filtering [J]. Journal of Computer Applications, 2014, 34(3): 915-920.
[10]	CHEN Zhifang WANG Jinglei SUN Jingsheng LIU Zhugui SONG Ni GAO Yang. Multi-source irrigation information fusion method based on fuzzy rough set and D-S evidence theory [J]. Journal of Computer Applications, 2013, 33(10): 2811-2814.
[11]	LIU Hai-yan YANG Chang-yu LIU Chun-ling ZHANG Jin. Tracking algorithm for moving objects based on gradient and color [J]. Journal of Computer Applications, 2012, 32(05): 1265-1268.
[12]	. Design and implementation of fall detection system using tri-axis accelerometer [J]. Journal of Computer Applications, 2012, 32(05): 1450-1452.
[13]	. Electronic image stabilization algorithm based on filtering and curve fitting [J]. Journal of Computer Applications, 2010, 30(11): 3008-3010.
[14]	Wen-Qiang GUO Zhi-Guang QIN. Research on RBFNNaided adaptive UKF algorithm [J]. Journal of Computer Applications, 2009, 29(3): 858-861.
[15]	Guo-qiang Xiao . Kalman filtering for tracking video objects based on motion estimation [J]. Journal of Computer Applications, 2008, 28(8): 2052-2054.