面向海量高维数据的DEA效率评价理论、方法与应用

数据包络分析（DEA）是组织效率评价的重要工具。在大数据时代，数据规模和数据维度呈爆炸式增长，不可避免地导致DEA计算能力受限和维数诅咒的双重难题，这给DEA效率评价带来严峻挑战。因此，本项目拟开展面向海量高维数据的DEA效率评价理论、方法与应用研究。首先，根据海量高维数据的特征和DEA的效率评价机制，研究建立海量高维数据DEA的效率计算框架、估计量统计性质、结果质量评判标准以及效率评价的影响机理等基本理论。其次，分别提出单产出情形、多产出情形和数据不确定情形下的效率评价方法，构建海量高维数据DEA的效率评价方法体系。最后，进一步研究前述理论和方法在我国城市环境效率评价中的应用，验证所提出理论和方法的有效性。本项目的研究有助于推动大数据DEA理论的发展，并为我国城市环境管理提供重要的方法和决策支持。

Data envelopment analysis (DEA) is an important tool for organizations’ efficiency evaluation. In the big data era, the amount and dimensions of data increase dramatically, which inevitably leads to two problems of limited DEA computing power and the curse of dimensionality. And it brings severe challenges to DEA efficiency evaluation. As a result, this project will study the theory, methodology and application on DEA efficiency evaluation for big-and-wide data. First, according to the characteristics of big-and-wide data and the efficiency evaluation mechanism of DEA, we will examine and build the basic theory of calculation framework, statistical properties and quality evaluation standard of DEA efficiency, and the impact mechanism of big-and-wide characteristics on efficiency evaluation. Second, the efficiency evaluation methods will be proposed under the circumstance of single output, multiple outputs and uncertain data, respectively. And the DEA methodology on big-and-wide data will be developed. Finally, we will further investigate and apply the previous theory and methods to environmental efficiency evaluation of Chinese cities to verify the effectiveness of the proposed theory and methods. This project will promote the theory development of big data DEA, and provide important method and decision supports for city environmental management in China.