径向基神经网络在港城经济系统模型变量选取中的应用  

Application of RBF in choosing the port-city economic system model variables

作　　者：刘丽娜[1] 宋向群[1] 郭子坚[1] 

机构地区：[1]大连理工大学土木水利学院,辽宁大连116023

出　　处：《哈尔滨工程大学学报》2006年第B07期169-172,共4页Journal of Harbin Engineering University

基　　金：国家自然科学基金（50578030）.

摘　　要：针对构建港口城市经济系统（以下简称港城经济系统）模型中所遇到的问题，提出了基于径向神经网络的解决方法．即首先通过专家问卷调查，选取港城经济系统中定性分析的普遍影响因素，并以大连市为例，对不可量化因素进行数量化处理，建立径向基神经网络模型．通过历史数据对此模型进行训练和检验之后，利用该模型对港城经济系统的各个因素进行了敏感性分析，定量确定了大连市港城经济系统的主要影响变量，为后续的港城经济系统分析与研究提供了支持．A method is extracted to solve the problems that exist in the research of port-city economic system done in past. First of all, the influencing factors of port-city economic system have been chosen by expert consulting and the model of RBF in Dalian City has been developed based on quantifying the nonquantifiable factors. After training and checking the model using data collected, the sensitivity analysis of factors in port-city economic system has been made in the model. Through the sensitivity analysis, the major influencing factors of port-city economic system in Dalian have been determined, which supplied support for the subsequent analysis and research of the port-city economic system.

关 键 词：港城经济系统 专家问卷调查 数量化 径向基神经网络 敏感性分析 

分 类 号：F293[经济管理—国民经济]