基于数据挖掘的纺织业成本预警研究

发布时间：2018-02-28 06:27

本文关键词： 纺织业成本预警数据挖掘 BP神经网络支持向量机 AC算法　出处：《福州大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着现代工业飞速发展,市场竞争日益激烈,成本已成为影响行业发展的重要因素,所以成本的管理就变得非常关键的。纺织业作为我国国民经济传统支柱产业,为社会发展做出了很大的贡献。2012年,为了贯彻落实《国民经济和社会发展第十二个五年规划纲要》和《工业转型升级规划(2011-2015年)》,工业和信息化部制定并发布了《纺织工业“十二五”发展规划》,指出了纺织行业在国际和国内上都面临巨大的挑战,汇率的变动、国际间贸易壁垒、行业原材料价格上涨、劳动成本增加、技术改革、环境、能源等方面的因素都很大地影响纺织行业的发展,而且各因素都间接或直接地影响着行业的成本发生,对于一个利润空间本来就小的纺织行业,成本的管理就变得尤为重要。因此,研究纺织业的成本,建立行业的成本预警系统是有必要的。凡事预则立,不预则废,建立有效的成本预警系统可以帮助行业进行成本分析,控制成本风险,提高行业经济效益,进而促进纺织业的可持续发展。所以,本文立足于纺织业,进行成本预警研究。首先,阐述了进行纺织业成本预警研究的背景及意义,总结国内外相关研究现状；并介绍了成本预警相关理论、数据挖掘的定义、BP神经网络算法、支持向量机算法和相似体合成算法。其次,根据成本预警相关文献研究和纺织业现状,从定性的角度分析影响成本发生的因素,再结合灰关联法的定量分析,确立纺织业成本预警指标体系；划分警情指标的警度值,量化预警区间；并利用逐步回归法进行指标约简,消除不显著变量,减少模型的输入维数；然后基于数据挖掘的方法构建BP-SVM-AC预警模型,分两个阶段进行成本预警研究,先利用一阶段改进的BP-SVM模型确定成本警情的警度和强度,再借助AC算法,进行二阶段的成本警情预测。最后,借助数值计算软件Matlab R2012a进行实证研究。通过对BP-SVM-AC预警模型的结果分析,证明了该模型具有较好的预警效果。所以本文建立的BP-SVM-AC预警模型具有一定的科学性和实用性,并且通过该模型可以预测纺织业未来成本警情,帮助该行业的相关管理部门提前掌握成本警情的发展趋势,有利于他们及时采取措施应对警情发生,进而也为行业企业进行成本预警提供一定的参考价值。
[Abstract]:With the rapid development of modern industry and the increasingly fierce market competition, cost has become an important factor affecting the development of the industry, so cost management has become very critical. Made a great contribution to social development. In 2012, In order to implement the outline of the Twelfth Five-Year Plan for National Economic and Social Development and the Plan for Industrial Transformation and upgrading 2011-2015, the Ministry of Industry and Informatization formulated and issued the 12th Five-Year Plan for Textile Industry, pointing out that. The textile industry is facing enormous challenges both internationally and domestically. Changes in exchange rates, international trade barriers, rising prices of raw materials in the industry, increased labour costs, technological reforms, environment, energy and other factors have greatly affected the development of the textile industry. And all factors have an indirect or direct impact on the cost of the industry, and for a textile industry where profit margins are small, cost management becomes particularly important. It is necessary to set up an industry cost warning system. It is necessary to establish an effective cost warning system that can help the industry to carry out cost analysis, control the cost risk, and improve the economic benefits of the industry. Therefore, this paper bases on the textile industry, carries on the cost early warning research. First, elaborated the textile industry cost early warning research background and the significance, summarized the domestic and foreign correlation research present situation; The related theory of cost early warning, the definition of data mining, the BP neural network algorithm, the support vector machine algorithm and the similar volume synthesis algorithm are introduced. Based on the qualitative analysis of the factors affecting the occurrence of cost, combining with the quantitative analysis of grey correlation method, the paper establishes the cost early warning index system of textile industry, classifies the alarm value of the warning index, and quantifies the early warning interval. Using stepwise regression method to reduce the index, eliminate the unsignificant variables, reduce the input dimension of the model, then based on the method of data mining to construct the BP-SVM-AC early warning model, and carry out the cost early warning research in two stages. First, the one-stage improved BP-SVM model is used to determine the alarm degree and intensity of the cost alarm, and then the two-stage cost alarm prediction is carried out with the help of AC algorithm. Finally, Through the analysis of the results of the BP-SVM-AC early warning model, it is proved that the model has good early warning effect. Therefore, the BP-SVM-AC early-warning model established in this paper is scientific and practical. And through this model, we can predict the future cost warning of textile industry, help the relevant management department of this industry to grasp the development trend of cost warning in advance, and help them to take measures to deal with the situation in time. Furthermore, it also provides certain reference value for industry enterprises to carry out cost warning.
【学位授予单位】：福州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F426.81;F406.7;TP311.13

【相似文献】