基于支持向量回归机的耕地保有量组合预测

为提高耕地保有量预测精度,将灰色预测GM(1,1)模型、动力预测模型、BP网络预测模型和加权支持向量回归机预测模型相结合,建立了基于支持向量回归机的耕地保有量组合预测模型,并将其用于温州市耕地保有量预测。结果表明,该模型比任一单一预测模型精度更高,可用于耕地保有量预测。

Combination Forecasting for Total Farmland Based on Support Vector Regression

Based on the combination of grey forecasting model GM(1,1),dynamic forecasting model,BP neural networks forecasting model and weighted support vector regression forecasting model,a support vector regression combination forecasting model for total farmland forecasting is proposed to improve accuracy of total farmland forecasting.This model is tested on the total farmland of Wenzhou City,Zhejiang,and the results indicate that the combination forecasting method is of better accuracy than others.It is concluded that the method is feasible and effective in total farmland forecasting.