关键词: 生产前沿面, 核方法, 回归, 生产决策单元, 参数模型

Abstract: In order to solve the problem that it is difficult to select the production function in parametric production frontier analysis，a single-side support vector regression model for parametric production frontier is proposed.In the proposed kernel method based model，the input resources data of all decision making units in data space is mapped into feature space，thus the data can be linearly used.In this way，the nonlinear mapping of linear production function is employed to solve the problem of selecting a production function.At last，an experiment of evaluating economic efficiencies of cities in Pearl River Delta is conducted to verify the proposed model.

Key words: production frontier, kernel methods, regression, decision making unit, parametric model