基于支持向量顺序回归机对农业经济预警的研究

Study on Agricultural Economics Early-warning Based on Support Vectorordinal Regression Machines

在统计学习理论与支持向量机理论方法的基础上,将支持向量顺序回归算法与宏观经济预警研究相结合,建立了基于支持向量顺序回归经济预警模型;通过编制Matlab软件程序,结合黑龙江省经济数据,获得经济预警的分类超平面和预测警限区间,并进行内插和外推检验。结果表明,该预警模型预测的可靠性和拟合是有效可行的,具有很高的精度。

On the basis of statistical learning theory and support Vector Ordinal Regression Machines theory,combine the method of support Vector Ordinal Regression and the study of macroscopical economy early-warning.Establish economics early-warning model based on Support Vector Ordinal Regression.Through generating Matlab software programme,integrating economic data in Heilongjiang province,attaining economic early-warning classification ranges from above-level to forecasting warning line.Testing interpolation and extrapolation.The results show the credibility of this early-warning model and effective fitting with high precision.