The wuchuan county locating the ecological vulnerable region of northern foot of yin mountain was studied and divided by the principal components cluster into the four ecological economical regions of returning farmland to forest land , southeast mountain agricultural region , north hillside agricultural region , northwest wind erosion and sandy region of joining agriculture and herbage , and southwest region of joining agriculture and herbage 本文以位于阴山北麓生态脆弱带的武川县为研究对象，通过主成份聚类分析将武川县划分为东南部山地农业区、北部丘陵农业区、西北部风蚀沙化农牧结合区和西南部山地农牧结合区等四个退耕还林生态经济类型区。
Starting from the intention and principle of agriculture sustainable development , this article delves further into the frontier and hotspot issue - measurement mean namely the indicators system and evaluation pattern of sustainable development , quote the assembled evaluation method of multi - indicator synthesis evaluation theory into the evaluation of sustainable development , and then put forward a conclusion that the principle component - analytic hierarchy process analysis is much better than other methods 文章从农业可持续发展的内涵及原则出发，对农业可持续发展理论的前沿及热点问题? ?可持续发展的度量手段即指标体系及评价模式进行深入探讨，将多指标综合评价理论中的组合评价方法引用到可持续发展评价中，提出主成份? ?层次分析组合评价法优于其它方法的结论。
It concentrates east of shan dong , and it is relevant to level of economy . the paper adopt principal components method of factor analysis to study influence factors of human capital and region distribution feature . according to factor score , it adopt cluster analysis to class human capital different type which influencing the sustainable development of regional economy and its distribution 山东省人力资本在17地市分布差异明显；偏集于东部地区，与经济发展水平有高度相关性；采用主成份型因子分析方法分析影响人力资本的因素及其区域分布特征，根据各地区的因子得分，采用聚类分析的ward法很音差平方和法）进行聚类，划分山东省推动经济可持续发展的人力资本区域类型。
The best qualitative results indicated that rvpcs model with md grouping scheme reached a discrimination rate of 100 % with 1 % accuracy in the randomly assigned background constituents of the same prescription , and rvpcs model with rs grouping scheme gave a discrimination rate of 86 % ~ 96 % with 2 % accuracy in the randomly assigned background constituents of 4 mixed prescriptions 四种处方之不同背景混合处方分析中，以主成份空间之残差变异定性模式搭配随机分组方式，可达到86 % ~ 96 %之辨识成功率，并具有2 %之微量检测精度。
