多变量混沌时序正则化局部线性预测

Regularized Local Linear Prediction of Multivariable Chaotic Time Series

为了克服多变量混沌时序局部线性预测模型中利用最小二乘法确定参数时会产生多重共线性的缺陷,提出了基于正则化回归的多变量混沌时序局部线性预测模型。该预测模型是在一般的多变量混沌时序局部线性预测模型中对最小二乘法进行改进,引入正则化估计,利用正则化回归法对模型参数进行估计。实证研究结果显示,模型比改进前有更好的预测精度和抗噪能力。

In order to overcome the shortcoming of multicollinearity when using the model of least square in the local linear prediction model of the multivariate chaotic time series, a local linear prediction model of multivariate chaotic time series based on the regularized regression is put forward. The advanced model is to improve the least square method by introducing regularized estimators in the generalized multivariate local lineal prediction model, and to estimate the parameter in the model with regularized regression. The results show that the model has better prediction precision and noise reduction.