摘要
基于农业资源利用效率评价指标的多目标性和不相容性,提出了农业资源利用效率综合评价的遗传投影寻踪方法。该方法可以依据样本自身的数据特性寻求最佳投影方向,从而判断各评价指标对综合评价目标的贡献大小和方向。通过最佳投影方向与评价指标的线性投影得到投影指标值,通过这一指标可以对样本进行统一评价。利用该方法对甘肃省81个县域单元的农业资源利用效率进行综合评价,并对综合评价结果进行分级。
A new method, project pursuit based on genetic algorithm, was used to evaluate agricultural resources use efficiency for solving the multi - objective and incompatibility of its evaluation indexes. Project pursuit can be used to project high dimensional data to low dimensional space and find the optimum projection vector of data in one dimension space by data character of research units, which can comprehensively evaluate the value and direction of every index. The projection index values were gave by multiplying projection vector by standardization value of evaluation indexes, which can comprehensively evaluate agricultural resources use efficiency of every unit. The method was used to synthetically evaluate agricultural resources use efficiency of 81 counties in Gansu province, and the results were classified.
出处
《经济地理》
CSSCI
北大核心
2006年第4期632-635,共4页
Economic Geography
基金
中国科学院知识创新工程重要方向项目(编号:KZCX3-SW-333)资助。
关键词
农业资源利用效率
综合评价
投影寻踪方法
遗传算法
agricultural resources use efficiency
comprehensive evaluation
projection pursuit
genetic algorithm
