基于支持向量回归与地统计学的农民收入预测

农民收入既受其自身动态时序特性的影响,又与国家政策、农产品产量、天气等多种人为因素和自然因素关系密切,是一种典型的多维时间序列数据.该研究以国民总收入、乡村人口等11个可能与农民收入相关的影响因子为自变量,农村居民家庭平均每人纯收入为因变量,基于1982～2011年我国相关统计数据,采用支持向量回归与地统计学非线性时间序列预测模型对农民收入进行分析.结果表明：非线性时间序列预测模型大幅度提高了农民收入预测精度;农民的主要收入来源已不是粮食生产,而是向其他农业生产、进城务工等方面转移.

Prediction of Famers' Income Based on Support Vector Regression and Geostatistics

Farmers＇ income is affected by its own dynamic temporal features, and is closely related to human factors and natural factors such as the national policy, agricultural production, weather and so on, which is a kind of typical multidimensional time series data. Using the gross national income, rural population and so on 11 may influence the farmers＇ income factors as the dependent variable, farmers＇ income as the target variables, based on the data of 1982 -2011, the SVM -GS-fiher was used to analyze farmers＇ income. The results showed that the main source of farmers＇ income has not been the usual grain production, but transferred to other agricultural production, migrant workers etc.