GMDH网络在进出口贸易预测中的应用

随着世界经济一体化的加快,对外进出口贸易在一国经济中发挥越来越重要的作用。进出口贸易影响因素较多,波动性大,一般方法难以准确预测。针对该问题,采用基于PSO优化的GMDH网络预测出口贸易总额,该方法首先根据输入数据特点自适应构造GMDH网络,然后用PSO算法优化GMDH网络权值,最终得到较为精确的预测器。通过仿真结果表明,该方法能够较为准确的预测中国年度进出口贸易总额。

Application of the GMDH network prediction in the import and export trade

With the development of world economy integration, import and export trade plays a more andmore important role in a nation＇ s economy. But import and export trade is influenced by many factors andis volatile. According to the problem, the article proposed a GMDH network import and export predictionmethod based on PSO. The method adaptively constructs GMDH network according to the data, then PSOis employed to optimize GMDH network weights in order to predict accurately. The simulation resultsshow that the method can predict the total volume of Chinese annual import and export trade.