基于商业智能的制造业上市公司财务预警研究-企业管理专业论文.docxVIP

摘 摘 要 I万方数据 I 万方数据 摘 要 本文在回顾了国内外关于财务预警和商业智能的相关文献，选取了 2010-2013 年 期间首次被 ST 的 50 家制造业上市公司和与其相配对的 50 家非 ST 上市公司，以首 次被 ST 的 T-3 年的财务数据对制造业上市公司的财务预警进行研究。首先本文根据 指标选取原则选取了 29 个涉及盈利能力、偿债能力以及营运能力等五个方面的财务 指标，并通过了正态性检验、T 检验和非参数检验筛选出了 24 个显著性指标构建了 财务预警指标体系；其次利用商业智能软件 SQL server 2008 建立了数据仓库，并通 过相关的查询语句实现财务数据的多维联机分析和数据透视表的呈现。最后通过 SSAS 数据挖掘模块的决策树、聚类分析、逻辑回归和神经网络算法构建了 4 个预警 模型，汇总了预警模型中较为重要的指标，并使用提升图和分类矩阵对模型的有效性 进行比较分析，提取出隐藏在样本中的潜在信息。本文创新性的使用 SQL server 2008 商业智能软件构建了财务预警体系，通过整合数据源创建数据库达到了数据的及时传 输，完善了财务的实时监控功能，并有效的分析了生成的财务预警模型，实现了会计 信息智能化发展，运用了模型的比较分析，选取了制造业最优模型。 结论表明：营业利润率，流动比率和利息支出倍数这几个比率对财务危机的预警 比较重要；决策树算法建立的挖掘模型可以对制造业上市公司未来是否发生财务危机 进行较为正确的判断，预测的准确性达到 80%。 关键词：财务预警 商业智能 实时监控 Abs Abstract Ⅱ万方数据 Ⅱ 万方数据 Abstract By reviewing many domestic and international financial early warning and business intelligence in classic literatures, the sample of study are new added 50 ST manufacture companies from2010 to2013 which are compared with other 50 well-run companies in the same sub-industry, and researches the manufacture listed companies’ financial crisis early warning by the financial data of three years earlier by the time the companies was specially treated. First, according to the index selection principle, this article selected the 29 financial indicators involving five aspects of the profitability, solvency and operation capacity and so on. Based on the normality test, T test and nonparametric test, 24 indexes was considered to have significant difference, and build the financial early warning index system. Then financial data warehouse was established by BI software SQL server 2008, which can achieve the OLAP and present the Pivot Table by applying the related query. Finally, we get the 4 financial early warning models with the help of four algorithm of decision tree, cluster analysis, logistic regression and neural network in data mining modules of SSAS, by summarizing the more important indicators of the early warning model and using the lift craft and classification matrix to anal