基于DEA和神经网络集成模型的我国基础设施投资有效性预测研究

运筹与管理 ›› 2011, Vol. 20 ›› Issue (6): 88-98.

基于DEA和神经网络集成模型的我国基础设施投资有效性预测研究

李玉龙^1,2, 李忠富²

1.中央财经大学管理科学与工程学院, 北京 100081;
2.哈尔滨工业大学管理学院, 黑龙江哈尔滨 150001

收稿日期:2009-12-05 出版日期:2011-12-25
作者简介:李玉龙(1980-),男,黑龙江绥化人,博士,讲师,研究方向为决策预测评价、经济系统工程;李忠富(1964-),男,黑龙江呼兰人,教授,博士生导师,研究方向为建设经济、城市可持续发展。
基金资助:
国家自然科学基金资助项目(G0724003);中财121人才工程青年博士发展基金资助项目(QBJGL201006)

Combined DEA and Neural Network for Predicting Investment Validity of Infrastructure on China

LI Yu-long^1,2, LI Zhong-fu²

1. School of Management Science and Engineering, Central University of Finance and Economics, Beijing 100081, China;
2. School of Management, Harbin Institute of Technology, Harbin 150001, China

Received:2009-12-05 Online:2011-12-25

摘要/Abstract

摘要： 建立了DEA和神经网络集成的基础设施投资有效性预测模型。该模型首先应用DEA方法,对我国1993-2007年逐期的基础设施投资效率进行评价,得到了用于基础设施投资有效性预测的基本数据。根据对评价结果的投资有效和无效划分建立预测样本,选择多层感知器神经网络,分别对基础设施的规模有效性和技术有效性进行了预测。结果表明基础设施的投资有效性预测具有可行性,而且通过与RBF神经网络、logistic回归和C-支持向量分类机等方法对比,MLP-NN方法的回应率和反查都具有优势,表明应用DEA-MLP-NN进行有效性预测更为有效。

关键词: 工程管理, 预测, 数据包络分析, 神经网络, 基础设施投资

Abstract: Combined model of data envelopment analysis(DEA)and neural network for predicting investment validity of infrastructure on China is proposed in this paper. Firstly, investment efficiency on infrastructure based on DEA method from 1993 to 2007 is evaluated to obtain the basic data to predict investment validity. And then, according to the classifying samples which is established based on the evaluated results with DEA method, the scale validity and technical validity of infrastructure is separately predicted with the multi-layer perceptron neural network (MLP-NN). The results show that the prediction of investment validity on infrastructure is feasible, and the response rate and the recall have an obvious advantage by comparing with RBF neural network approach and C-SVM method and logistic regression. DEA-MLP-NN method is more effective.

Key words: engineering management, predicting, DEA, neural network, infrastructure investment

中图分类号:

F201

李玉龙, 李忠富. 基于DEA和神经网络集成模型的我国基础设施投资有效性预测研究[J]. 运筹与管理, 2011, 20(6): 88-98.

LI Yu-long, LI Zhong-fu. Combined DEA and Neural Network for Predicting Investment Validity of Infrastructure on China[J]. Operations Research and Management Science, 2011, 20(6): 88-98.

参考文献

[1] 陈赟.论提高政府投资效益[J]. 重庆大学学报:社会科学版, 2006,12(1):40-45.
[2] Wu D S, Yang Z J, Liang L. Using DEA-neural network approach to evaluate branch efficiency of a large canadian bank[J]. Expert Systems with Applications, 2006, 31(1): 108-115.
[3] Mostafa M M. Modeling the efficiency of top Arab banks: A DEA-neural network approach[J]. Expert Systems with Applications, 2009, 36(1): 309-320.
[4] Tsai M C, Lin S P, Cheng C C, et al. The consumer loan default predicting model-an application of DEA–DA and neuralnetwork[J]. Expert Systems with Applications, 2009, 36(9): 11682-11690.
[5] Wu D S. Supplier selection: a hybrid model using DEA, decision tree and neural network[J]. Expert Systems with Applications, 2009, 36 (5): 9105-9112.
[6] Wang C H, Chuang C C, Tsai C C. A fuzzy DEA-neural approach to measuring design service performance in CM projects[J]. Automation in Construction, 2009, 18(5): 702-713.
[7] Emrouznejad A, Shale E. A combined neural network and DEA for measuring efficiency of large scale datasets[J]. Computers & Industrial Engineering, 2009, 56(1): 249-254.
[8] Aschauer D. Is public expenditure productive[J]. Journal of Monetary Economics, 1989, 2(23): 177-200.
[9] Stephan A. Essays on the contribution of public infrastructure to private production and its political economy[C]. Humboldt University, 2001. 9-12.
[10] 王任飞,王进杰.基础设施与中国经济增长:基于VAR方法的研究[J].世界经济,2007,(3):13-21.
[11] 魏权龄.数据包络分析[M].北京:科学出版社,2004,8:2-20.
[12] Mastromarco C, Woitek U. Public infrastructure investment and efficiency in italian regions[J]. Journal of ProductivityAnalysis, 2006, 25(4): 57-65.
[13] Fedderke J W. Infrastructural investment in long-run economic growth: south africa 1875-2001[J]. World Development, 2006, 34( 6): 1037-1059.
[14] 李世蓉,蒋时节,户邑.城市化进程对基础设施投资的需求量分析[J].城市发展研究,2005,12(4):29-32.
[15] Rudra P P. Does infrastructure play role in urbanization: evidence from india[J]. Indian Journal of Economics and Business, 2007, (6): 113-124.
[16] 李忠富,李玉龙.基于DEA方法的我国基础设施投资绩效评价:2003-2007年实证分析[J].系统管理学报,2009,18(3):309-315.
[17] 韩力群.人工神经网络理论设计和应用[M].北京:化学工业出版社,2002.23-31,55.
[18] 庞素琳,王燕鸣.多层感知器信用评价模型研究[J].中山大学学报(自然科学版),2003,42(4):118.
[19] 谢进,阎开印,陈永.基于BP神经网络的平面机构连杆运动综合[J].机械工程学报,2005,41(2):26.
[20] 范秋芳.基于BP神经网络的中国石油安全预警研究[J].运筹与管理,2007,16(5):100-105.
[21] 杨淑娥,王乐平.基于BP神经网络和面板数据的上市公司财务危机预警[J].系统工程理论与实践,2007,2:62-77.
[22] 吴方良,石仲,杨向晖,陈锐.基于BP神经网络与L-M算法的潜艇声纳自噪声预报[J].中国造船.2006, 47(3):45-50.