论文部分内容阅读
针对农业机械化效率存在的差异，提出应用 k 均值聚类算法对新疆各地州按农机化效率进行区划，并以各地州2014年面板数据为研究对象进行区划。对区划结果应用误差平方和（ Sum of Squares for Error ）、轮廊系数（ Silhouette Coefficient ）方法进行检验和定量分析，同时对结果根据地区农机化发展现状进行定性分析。结果表明：新疆各地州按其农机化效率分为4组（k＝4）：第1组昌吉、塔城、阿勒泰、巴州，第2组克拉玛依、博州，第3组哈密、伊犁、阿克苏、喀什，第4组乌鲁木齐、吐鲁番、克州、和田；组间差异最大，组内差异最小。该研究可为分类指导各地区农机化发展提供参考。“,”The regionalization approach was proposed based on k-means clustering algorithm .On the clustering experi-ments, the 14 regions was zoned 3, 4, 5, 6,7and 8 zone according to efficiency of agricultural mechanization of each re -gion.The SSE ( Sum of squares for error) and silhouette coefficient method was applying for validation the quality of zoning ( clustering ) .On the same time the qualitative analysis for clustering result was applying on the basis of current situation of agricultural mechanization on each region in Xinjiang .The results show that , according to each region ’ s effi-ciency of agricultural mechanization in 2014 , if the all regions in Xinjiang be zoned for 4 group ( k=4 ) , regions were in the same group(cluster)are more similar to each other than to those in other groups (cluster) ,meanwhile the better SSE and silhouette can be obtained as well .The first group included Changji , Tarbaghatay , Altay and Bazhou .The second group included Karmay , Bortala, The third group included Kumul , Yili, Ahsu and Kashghar .Urumqi, Turpan, Kezhou and Hotan were included in forth group .Our approach and results provide useful information for development of agricul -tural mechanization in management .