基于主成分回归模型的工程项目成本预测

发布时间：2018-03-08 22:40

本文选题：工程成本预测　切入点：主成分因子分析　出处：《扬州大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着市场经济的不断深化,建筑企业在获得更多发展机遇的同时也面临着更大的竞争压力。企业的竞争优势需要靠每个项目可观的经济效益来保证,那么科学的成本管理作为实现经济效益的手段就显得更为重要。工程成本预测做为工程成本管理工作中的先前节点,是体现一个企业成本管理水平的重要所在。在工程信息化的发展趋势下,已建工程的成本信息数据库可以为工程成本预测的准确性提供良好的基础。建筑企业可以充分挖掘这些数据的潜在信息并结合自身现有的管理能力,在项目投资决策前,进行科学有效的成本预测。为了适应工程数据信息化处理的发展要求,本文提出借助线性回归模型从大量工程成本信息数据的特征中分析影响因子间的线性关系,建立成本预测回归方程。并提出数据挖掘技术与线性回归模型结合使用,可以更准确的拟合因变量与自变量之间的相关关系,从数据中找出“相关趋势”,可以进行精度较高的中长期的成本预测,有利于管理者对项目成本进行事前控制。论文主要的研究内容如下： 1.收集大量已建工程的成本数据信息,根据工程成本造价的组成,从定性因素和定量因素的角度,初步筛选出工程成本预测的影响因子,并拟用线性回归建立工程成本预测模型。 2.因为利用SPSS数据统计分析软件初步筛选出的影响因子较多,针对影响因子之间可能会出现的多重共线性问题,提出使用主成分因子分析法来提取新的变量因子。主成分因子分析法的使用在达到降维目的的同时,也简化了线性回归模型的建立,在实际的工程成本预测中具有更好的可操作性。 3.根据提取出的新的变量因子建立主成分回归预测模型,并对预测模型进行检验。通过实际算例,验证了本文中已建模型的可靠性。 4.对文中的主成分回归预测模型进行分析,总结了论文研究的结果与不足,为了今后工程成本预测的进一步研究提出展望。
[Abstract]:With the deepening of market economy, construction enterprises are facing greater competition pressure to get more development opportunities at the same time. The competitive advantage of enterprises need to rely on each project to ensure considerable economic benefits, so the cost of management science as a means to achieve economic efficiency becomes more and more important. As the previous node engineering cost management in the work of project cost forecast, cost management is an important embodiment of the level of an enterprise.
The development trend of engineering information, cost information database project provides a good basis for the accuracy of project cost forecast. Construction enterprises can fully tap the potential information of these data combined with its existing management ability in the project investment decision, prediction of scientific and effective cost. In order to adapt to the development of information engineering data the processing requirements, this paper propose using linear regression model to analyze the effects of linear relations among the factors from the characteristics of a large number of engineering cost information data, establish the cost prediction regression equation. And put forward the technology of data mining and the linear regression model is used in combination, can be more accurate fitting for the relationship between variables and independent variables, find out the "trend from the data, can be predicted with high accuracy in the long-term cost, is conducive to the management of the project cost. Control in advance.
The main contents of this paper are as follows:
1., collecting a large number of cost data information of the existing projects, according to the composition of the cost of the project, from the perspective of qualitative factors and quantitative factors, we preliminarily screened out the impact factors of the project cost prediction, and set up the prediction model of the project cost by linear regression.
2. because there are many factors that affect the use of SPSS statistical analysis software was screened for multiple, likely to affect the factor between the multicollinearity problem, proposed using principal component factor to extract new variable factor analysis. Principal component analysis using dimensionality reduction to achieve at the same time, but also simplifies the establishment of linear regression the model, with better operability in predicting the actual project cost.
3., we build a principal component regression prediction model based on the extracted new variables, and test the prediction model. The reliability of the model built in this paper is verified by a practical example.
4., the principal component regression prediction model is analyzed, and the results and shortcomings of the research are summarized. The prospect for further research of engineering cost prediction is put forward.

【学位授予单位】：扬州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU723.3

【参考文献】