期刊文献+

基于多层前馈网络的计量经济模型敏感性分析方法 被引量:1

Sensitivity Analysis of Econometric Model Using Multilayer Feedforward Network
下载PDF
导出
摘要 多层前馈网络计量经济模型可以描述复杂的非线性经济现象,在经济建模中的应用曰益广泛。而经济敏感性分析是一种研究经济的有效方法,潜在函数关系的可微性是应用该方法的前提条件之一,其要点是用拟合函数的导数来近似潜在函数的导数。介绍了四种敏感性分析方法,给出了多层前馈网络计量经济模型正规化灵敏度的解析计算方法,并说明了对多层前馈网络计量经济模型进行敏感性分析是一种有效识别敏感因素的方法。 Multilayer feedforward network can describe the nonlinear economic phenomena of various complexities which has been widely used to perform the economic modeling. Economic sensitivity analysis is an efficient method to study the economic systems, which requires the differentiabiity of the underlying function. The essential of this technique is to use the derivatives of the fitted function to approximate the derivatives of the underlying function. Four methods of sensitivity analysis were presented, and the analytical computation of the normalized sensitivity of the econometric model using multilayer feedforward network was given. The applicatios have shown that sensitivity analysis to this model is an efficient tool for the identification of the sensitive factors.
出处 《科学技术与工程》 2004年第9期786-788,共3页 Science Technology and Engineering
基金 国家自然科学基金(50244015)资助
关键词 多层前馈网络 计量经济模型 敏感性分析 正规化灵敏度 multilayer feedforward network econometric model sensitivity analysis normal- ized sensitivity
  • 相关文献

参考文献6

  • 1田野,孙永广,吴宗鑫.基于神经网络计量经济学模型经济扰动分析的研究[J].数量经济技术经济研究,2002,19(8):94-97. 被引量:3
  • 2[2]Fisher R A. The use of multiple measurements in taxonomic problems. Annual Eugenics, 1936 ;7(2): 179-188
  • 3[3]Cybenko G. Approximation by superposition of a sigmoidal function. Mathematics of Control Signals and Systems, 1989; 2 ( 4 ):303-314
  • 4[4]Hornik K, Stinchcombe M, White H. 0niversal approximation of an unknown mapping and its derivatives using multilayer feedforward networks. IEEE Transactions on Neural Networks,1990;3(5) :551-560
  • 5[5]Hashem S. Sensitivity analysis for feedforward artificial neural networks with differentiable activation funcitons. Proceedings of the 1992 IEEE International Joint Conference on Neural Networks, 1992;1:419-424
  • 6[6]Gallant A R, White H. On learning the derivative of an unknown mapping with multilayer feedforward netwroks. IEEE Transactions on Neural Networks, 1992;5:129-138

共引文献2

同被引文献5

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部