高新技术园区投入产出绩效的智能评价  

Intelligent Evaluation of Input and Output Performance for High-Tech Industrial Zones

作　　者：邢爱国[1] 孙佰清[2] 

机构地区：[1]哈尔滨工程大学经济管理学院,哈尔滨150001 [2]哈尔滨工业大学管理学院,哈尔滨150001

出　　处：《哈尔滨工业大学学报（社会科学版）》2006年第5期78-81,共4页Journal of Harbin Institute of Technology(Social Sciences Edition)

基　　金：哈尔滨市科技攻关项目(2004AA1CG132-15);2003年黑龙江省新世纪高等教育教学改革工程项目;黑龙江省青年科学技术专项资金项目(QC05C74);哈尔滨工业大学文科创新基金项目(050412)

摘　　要：投入产出绩效评价是关系着中国各高新技术园区可持续发展的关键问题。由于评价问题的高维复杂性,神经网络方法与其他方法相比具有一定的优越性,但存在着训练准则和分类准则不一致的问题。采用一种区域映射神经网络模型,实现了从区域到区域的映射比区域到点的映射更快、更好,保证了神经网络训练准则与实际分类准则的一致性,具有比前馈式神经网络更快的收敛速度和更高的分类精度。利用区域映射神经网络模型对中国高新技术园区投入产出绩效进行智能评价,取得了较好的效果。Evaluation of the input and output performance is the key problem which is concerned with the sustainable development of state high - tech industrial zones , in which due to the great complexity of evaluation, neural network has the disadvantage that its training rule is not consistent with classification rule, though it proves to be more valuable than other methods. This paper designs a region mapping neural network model, which guarantees the consistency of classification rule with the training rule by replacing the mapping of region to point by the mapping of region to region, and has faster convergence speed and better identification ratio than BP neural network. The model works well in intelligent evaluation of input and output performance for high - tech industrial zones.

关 键 词：高新技术园区 智能评价 区域映射神经网络模型 投入产出 

分 类 号：F270[经济管理—企业管理]