摘要
以2001年~2005年间的192家被认定为信息披露舞弊的A股上市公司及相应的192家配对公司为样本,基于财务指标和治理指标,分别运用Logistic回归分析和混合BP神经网络构建上市公司信息披露舞弊的预警模型。实证结果表明,治理指标有助于提高信息披露舞弊预警模型的有效性,混合BP神经网络模型的预测能力更强。
This paper collects a sample of 192 A-share listed companies committing disclosure fraud spanning from 2001 to 2005 and the corresponding 192 matched companies. Based on financial indicators and corporate governance indicators, the paper develops two disclosure fraud detection models by applying the Logistic regression and hybrid BP neural network method respectively, The empirical results show that corporate governance indicators help to improve the efficiency of the prediction model. Moreover, the hybrid BP neural network model dominates the Logistic regression model.
出处
《管理科学》
CSSCI
2006年第4期79-90,共12页
Journal of Management Science
基金
国家自然科学基金(70372035)