论文部分内容阅读
针对小样本非线性时间序列,根据非线性协整的定义,利用基于粒子群优化最小二乘支持向量机的方法,对小样本非线性协整关系检验与非线性误差修正模型建模进行研究,设计了方法的逻辑流程.对舰船维修费指数与物价指数进行实证研究,在协整关系类型判断的基础上,实现了小样本非线性协整关系的检验,建立了预测舰船维修费指数的非线性误差修正模型,并与线性向量自回归模型进行分析比较.研究表明:基于粒子群优化最小二乘支持向量机的小样本非线性协整检验与建模方法,刻画了小样本系统的非线性协整关系,所建立的非线性误差修正模型具有较好的预测效果,能够有效地预测小样本非线性系统.
According to the definition of nonlinear co-integration, a method based on Particle Swarm Optimization (LS-SVM) is proposed to study the nonlinear cointegration test and nonlinear error correction model modeling in small samples. The logic flow of the method is designed.An empirical study of the ship maintenance cost index and the price index is carried out. Based on the judgment of the type of the cointegration relationship, the nonlinear co-integration relationship of the small sample is tested and the prediction index of ship maintenance cost Linear error model and compared with linear vector autoregressive model.The results show that based on Particle Swarm Optimization (LS-SVM) small sample non-linear cointegration test and modeling method, the performance of the small sample system Nonlinear co-integration relationship, the established nonlinear error correction model has a good prediction effect, can effectively predict small sample nonlinear system.