基于灰色TOPSIS⁃CNN的上市企业财务绩效预测  

Prediction of financial performance of listed companies based on grey TOPSIS⁃CNN

作　　者：王可以 刘洪久[1] 胡彦蓉[1] Wang Keyi;Liu Hongjiu;Hu Yanrong(College of Mathematics and Computer Science,Zhejiang Agriculture and Forestry University,Hangzhou 311300,China)

机构地区：[1]浙江农林大学数学与计算机科学学院,杭州311300

出　　处：《现代计算机》2024年第18期15-21,共7页Modern Computer

基　　金：浙江农林大学学生科研训练项目资助(2023KX137)。

摘　　要：合理地评价和预测企业的财务绩效,对上市企业持续健康的发展非常重要。提出一种融合灰色TOPSIS和一维卷积神经网络的企业财务绩效预测模型。利用灰色TOPSIS对上市企业绩效做综合评价,通过K均值聚类生成标签,引入SMOTE算法解决数据不平衡问题,最后采用卷积神经网络进行绩效预测。结果显示,模型准确率达到96.35%,相比灰色TOPSIS⁃Kmeans⁃SMOTE⁃CNN、灰色TOPSIS⁃Kmeans⁃CNN、Kmeans⁃CNN、Kmeans⁃SVM、Kmeans⁃KNN平均提高5.76%,宏平均各评价指标也表现更优。表明本文提出的企业财务绩效分类模型性能较好,对企业财务绩效管理有积极的意义。Reasonably evaluating and predicting the financial performance of companies is very important for the sustainable and healthy development of listed companies.The paper proposes an company financial performance prediction model that integrates grey TOPSIS and one‑dimensional convolutional neural network.The grey TOPSIS is used to do comprehensive evaluation of the corporate performance of listed companies,the labels are generated by K‑mean clustering,the SMOTE algorithm is introduced to solve the problem of data imbalance,and finally the convolutional neural network is used for performance prediction.The results show that the accuracy of the model proposed in this paper reaches 96.35%,which is an average improvement of 5.76%compared with grey TOPSIS‑Kmeans‑SMOTE‑CNN,grey TOPSIS‑Kmeans‑CNN,Kmeans‑CNN,Kmeans‑SVM,and Kmeans‑KNN,and the macro‑averaged performance of each evaluation index is also better.It shows that our proposed corporate financial performance classification model performs better and has positive significance for corporate financial performance management.

关 键 词：卷积神经网络 灰关联 TOPSIS SMOTE 企业绩效 

分 类 号：F832.51[经济管理—金融学] F275