【作者】 翟坤；

【导师】 刘晓冰；

【作者基本信息】 大连理工大学 ， 企业管理， 2012， 博士

【摘要】 成本管理是管理领域的一个重要研究内容。目前的成本管理方法,大都是以财务管理方法角度,从解决策略上进行文字论述,以及采用基本的量化公式进行业务问题分析,缺少先进智能的定量化方法。同时,随着信息社会中数据的爆炸性增长,政府和企业等机构都积累了海量的成本数据资料。在知识经济新时代,如何将“数据的海洋”转变成“知识的金矿”,使得成本管理朝着数字化、智能化、知识化的方向发展,信息技术,尤其是数据挖掘技术的产生与发展为这一需求提供了强有力的支持。针对数据挖掘与成本管理的融合问题,本文在分析国内外数据挖掘技术、成本管理理论和两者相结合方法的研究及其应用现状基础上,以东北特钢集团实践为背景,应用多学科知识交叉融合来研究基于数据挖掘的成本管理方法的若干关键技术。首先,论文研究了基于数据挖掘的成本管理构成要素模型,介绍了其数据挖掘对象、成本数据来源、成本计算方法、以及各个组成要素,并阐述了如何在该模型下,实施对企业的数据挖掘任务,完成具体业务问题的解决方案。其次,提出了采用改进的关联规则算法来完成作业成本核算的作业选择与合并任务,实现了作业数据庞大规模情况下,高效地进行重要作业选择和重要作业合并；提出了基于优化模糊模型的成本预测方法,不仅对产品成本时间序列数据实现了较高精度的成本预测,而且使得预测过程具有类似于人类推理过程的运作机理,提高了成本预测方法的适应性；提出了基于改进粗糙集约简的成本决策因素识别方法,实现了相容成本决策表的规则生成最简化,获取了有效决策规则中精简优化后的成本决策因素；提出了基于优化动态模糊聚类的成本等级分析方法,实现了各核算期间的产品成本数据进行所属等级的自动划分,有效完成了动态的成本等级分析；提出了基于孤立点检测的例外成本控制方法,实现了产品成本数据的多维度情形下,发现例外偏离的产品成本及其显著偏离的具体维度,明确了例外成本控制的方向。最后,将论文研究成果与工程实际结合,以案例企业的成本管理需求、应用背景和实绩数据,来验证本文所述方法的有效性和可行性,研究成果为其他行业与企业实现智能化和知识化的成本管理提供了一定的参考与借鉴。更多还原

【Abstract】 Cost management is an important research content in management field. At present, cost management methods almost adopt character contents to describe some solving strategies, and use basic formulas to analyze some business problems from the perspective of financial management method. In that way, it lacks the advanced intelligent quantitative method. Meanwhile, with the explosive growth of data in the information society, government, enterprises and other institutions have accumulated vast amounts of cost data. In the new era of knowledge economy, how to transfer "data ocean" to "knowledge gold mine" and let cost management develop towards the digital, intelligent and knowledgeable direction, information technology, especially the emergence and development of data mining technology, can provide strong support for that requirements.In view of the integration issues between data mining and cost management, on the basis of analyzing the domestic and overseas research and application status of data mining technology, cost management theory and its combined method, as well as under the background of practice and application in Dongbei Special Steel Group Co. Ltd., a number of key technologies of cost management method based on data mining are studied by this paper through using multi-disciplinary knowledge and cross-application integration.Firstly, the paper studies the composing element model of cost management based on data mining, and introduces the model’s data mining object, cost data source, costing method and composing elements. Then how to implement the enterprises’data mining tasks for completing the solution programs of the specific business problem in the model is explained in detail. Secondly, the paper adopts an improved algorithm of association rule mining to accomplish the activity selection and mergence tasks of activity-based costing, which can efficiently realize the critical activity selection and critical activity mergence under the condition of large-scale activity data. Then, a cost prediction method based on an optimized fuzzy model is presented. It can not only realize the higher prediction accuracy for time-series data of product costs, but also make the prediction process resemble human reasoning process. It also improves the applicability of cost prediciton method. A factor identification method of cost decision based on an improved rough set reduction is also proposed, which can generate the minimization rules in the consistent cost decision table. It also get simplified and optimized cost decision factors in the effective decision rules. The paper also puts forward a cost rank analysis method based on an optimized dynamic fuzzy clustering algorithm. In each accounting period, it can automatically classify the product costs into the affiliated rank depending upon the structure of data themselves and the underlying complexity of data dynamics, which effectively realizes dynamic cost rank analysis. Moreover, an exceptional cost control method based on an outlier detection algorithm is also proposed. It can find out the exceptional product costs and its significant deviation dimension in the environment of multi-dimensional cost data, which clearly identifies the direction of exceptional cost control. Finally, by combining the thesis research and real-world practice, and following the case enterprise’s business requirements of cost management, application background as well as performance data, the results have demonstrated the effectiveness and practicability of the cost management method based on data mining. The research fruits can be cited and referred by other industries and enterprises for achieving their intelligent and knowledge-based cost management.更多还原