学生论文
|
论文查询结果 |
返回搜索 |
|
|
|
| 论文编号： | 961 | |
| 作者编号： | 2120071915 | |
| 上传时间： | 2009/5/26 10:39:00 | |
| 中文题目： | 基于粗糙集的企业财务失败预测研 | |
| 英文题目： | Corporate Financial Failure Pr | |
| 指导老师： | 安利平 | |
| 中文关键字： | 粗糙集 财务失败 比率分析 | |
| 英文关键字： | rough set financial failure | |
| 中文摘要： | 财务报告（或者财务报表）是企业披露财务状况的正式文档，财务分析师们利用财务报告对企业的财务状况进行分析，为企业、投资者、债权人和其他利益相关者提供建议。财务失败预测学派将财务报告分析的重心从历史评价转向未来预测，认为财务分析的重点是预测未来，特别是预测企业财务失败的可能性。粗糙集方法易于理解和应用，对数据集的统计性质要求不高，能够处理不完整和冗余数据，十分适用于财务数据的分析。 中国资本市场的迅速发展，尤其是新会计准则的实施为企业财务失败预测提出了新的要求，我们不能简单的继承西方预测财务失败的指标体系。本文深入分析中国资本市场现状，采用粗糙集的方法和理论进行财务比率分析，对财务失败预测进行分析、建模和应用。 本文遵循经典的CRISP-DM数据挖掘流程，综合利用SPSS、Clementine和Rosetta等数据挖掘软件，对粗糙集数据挖掘中的各种参数进行综合尝试，注重数据离散化和属性约简的方案。使用N次交叉验证的方法，力图从中找出最佳方案，最后将得到的结果联系经济环境进行综合分析。试验结果表明，文中的粗糙集模型同以往的同类研究相比，精确度有了一定程度的提高，而且结果验证了新会计准则对于财务分析的影响。 | |
| 英文摘要： | Financial reports (or financial statements) are the official documents for disclosing the financial position of businesses. Financial analysts use financial reports to analyze the financial situation of enterprises, providing businesses, investors, creditors and other stakeholders suggestions. The school of financial failure prediction shifts the focuses on historical evaluation to future predictions, regarding that the emphasis of financial evaluation is to predict the future, in particular, predicting the possibility of failure of corporate finance. Rough set theory provides the analysts of the prediction of financial failure a new tool. This method is easy to understand and use, requesting little for statistical nature of data sets, capable of handling incomplete and redundant data and analyzing financial data. The rapid development of China's capital market puts forward new demands for corporate financial failure prediction. Especially under the new accounting standards, we can not merely employ the Western system of indicators of predicting financial failure. This paper takes deep analysis of the status quo of China's capital market, using rough set methods and data mining theory, to conduct the financial failure analysis prediction, modeling and forecasting. In this paper, we follow the classic CRISP-DM data-mining process, with comprehensive utilization of SPSS, Clementine and Rosetta software. This paper tries various parameters, compares a lot of ways of discretization and reducts. Employing N-fold cross-validation and the stepwise method, we are trying to find the best option. Finally, the results are linked to a comprehensive analysis of the economic environments. Test results show that the accuracy of the rough set model has improved to some extent, compared to previous similar studies. Also, the results verify the new accounting standards’ impact on financial analysis. | |
| 查看全文： | 预览 下载(下载需要进行登录) |