广东耕地数量预测方法对比研究  

Comparison on the methods for predicting cultivated land quantity in Guangdong province

作　　者：吴明发[1,2] 欧名豪[2] 纪昌品[3] 郑建杰[1] 

机构地区：[1]广东海洋大学资源与环境系,广东湛江524088 [2]南京农业大学公共管理学院,江苏南京210095 [3]南昌大学公共管理学院,江西南昌330031

出　　处：《广东农业科学》2011年第19期207-210,共4页Guangdong Agricultural Sciences

基　　金：教育部博士点基金(20090097110022);湛江市科技攻关计划(2009C3112008)

摘　　要：以广东为例,运用GM(1,1)模型法、Holt双参数线性指数平滑法、多元线性回归模型法、平均增长法和时间序列移动平均法对其耕地数量进行预测,通过对预测结果残值的对比分析,发现移动平均法和GM(1,1)模型法具有较高的预测精度,并对各种预测方法进行了简要评价。Taken Guangdong province as an example,Gray Model （1,1）, Holt Curve Estimation and Exponential Smoothing, Multiple Linear Regression Model, Average Growth Method and Moving Average Method were employed to predict the cultivated land quantity in Guangdong Provinee. Through the comparative analysis of predictive residual, it was concluded that Moving Average Method and Gray Model （1,1） were better ways with higher accuracy. Finally, it also made a brief appraisal of the five methods.

关 键 词：耕地 预测 模型 广东 

分 类 号：F301.5[经济管理—产业经济]