摘要
针对财务信息正常与不正常样本数据集不平衡的问题,研究提出一种基于改进SVM的财务信息预警共享模型。首先,采用相关性检验法对财务信息初始特征进行检验,并确定了财务信息的重要特征;然后,采用改进的SVM算法构建了财务信息预警共享模型;最后,通过仿真实验验证了算法的可行性和有效性。结果表明,本研究提出的基于改进SVM算法的财务信息预警共享模型,对训练集样本分类正确率从91.46%,对测试集样本分类正确率为87.83%,相较于标准SVM算法和基于人工合成模型,本研究算法具有一定的优越性,可对财务信息进行准确分类。
Aiming at the problem of imbalance between normal and abnormal sample data sets of financial information,this paper proposes a financial information early warning sharing model based on Improved SVM.Firstly,the correlation test method is used to test the initial characteristics of financial information,and the important characteristics of financial information are determined.Secondly,the improved SVM algorithm is used to build the financial information early warning sharing model.Finally,the feasibility and effectiveness of the algorithm are verified by simulation experiments.The results show that the classification accuracy of the proposed model is 91.46%for the training samples and 87.83%for the test samples.Compared with the standard SVM algorithm and the synthetic model,the proposed model has certain advantages and can accurately classify the financial information.
作者
林海巍
李荣
Lin Haiwei;Li Rong(Beijing Shijitan Hospital,Capital Medical University,Beijing,100038;Beijing Easeek Technology Co.,Ltd,Beijing 100038)
出处
《现代科学仪器》
2022年第3期219-223,共5页
Modern Scientific Instruments
关键词
财务信息
数据分类
SVM算法
组合分类器
Financial information
data classification
SVM algorithm
combined classifier