基于支持向量机理论的经济圈交通网络评价研究

为快速高效地对经济圈交通网络进行评价分析,在确定经济圈交通网络评价指标后,采用新的评价方法——基于支持向量机的综合评价方法,通过计算机编程对样本数据学习,能较好地识别经济圈交通网络的评价等级。支持向量机对小样本、非线性的数据具有较好的学习识别能力,对具有相同数据特性的经济圈交通网络评价指标数据进行模式识别,可以准确地判断经济圈交通网络发展水平所属等级。最后对珠三角和长三角经济圈交通网络评价指标数据进行验证,并与基于BP神经网络的评价方法进行对比分析,结果表明该方法客观、准确,具有较高的评价效率。

A Study on Evaluation for Transportation Network in Economic Circle Based on Support Vector Machine Technology

The economic circle as a product of the development of social has closely relation with transportation.In this paper,a new evaluation method of small sample cases for transportation network is pre-sented based on support vector machine(SVM) technology,in view of the problems such as the relative complexity of evaluation methods,the strong non-correlation between indices,susceptibleness by the subjective factors and the weight of index no easy confirming and so on.Support Vector Machine is a new machine learning method based on the structural risk minimization principle,and has the high generalization performance.This method can solve the small sample,nonlinear and high-dimensional pattern recognition problems.So this method can be used to evaluate complex system and reject subjective interference factors,and it can be found in the samples to explore the internal law of the categories too.And the evaluation model of transportation network in economic circle is established by using the classification strategy of support vector machine.Finally,through simulation analysis for the sample of transportation network in the economic circle,the case of transportation network in the "Yangtze River Delta" and "Pearl River Delta" economic circle are evaluated and researched.The result of the case shows that the evaluation method has better classification results and can accurately reflect the level of economic circle transportation network development more objective.The algorithm of this method is more efficient too.