摘要
针对当前财务危机预测中高维数据训练耗时大、预测精度低的问题,提出一种基于自组织映射(SOM)聚类与宽度学习系统的财务危机预测方法。通过SOM对企业初始财务状态进行精细化聚类,提取出不同财务状态的指标特征;基于宽度学习系统的企业财务危机构建预测基分类器,确保对高维财务数据的拟合精度。利用逐步向前方法搭建集成分类器,预测企业财务状态。实例预测结果表明了该方法对财务危机预测具有较好效果,提升了预测准确度。
Aiming at the problems of high-dimensional data training time-consuming and low prediction accuracy in current financial crisis prediction, a financial crisis prediction method based on SOM clustering and width learning system is proposed. The initial financial status of enterprises is classified by SOM clustering. Based on the width learning system, the prediction base classifier is constructed to ensure the fitting accuracy of high-dimensional financial data. An ensemble classifier is built by the step-by-step forward method to predict the financial status of enterprises. The results show that the method has a good effect on financial crisis prediction and improves the accuracy of prediction.
作者
荘芳芳
ZHUANG Fangfang(Department of Finance,Kunming Technical School of Transportation,Kunming 650000,China)
出处
《微型电脑应用》
2022年第3期169-172,共4页
Microcomputer Applications
关键词
财务危机
SOM聚类
宽度学习
集成分类
financial crisis
SOM clustering
breadth learning
ensemble classification
