基于灰关联聚类降维数据空间的GM(1,n)预测模型  

GM(1,n)Prediction in Dimensionality Reduction Data Space Based on Grey Relational Clustering

作　　者：韩静 吴永武 郭本华 王辉 钱淑渠 杨正泽 HAN Jing;WU Yong-wu;GUO Ben-hua;WANG Hui;QIAN Shu-qu;YANG Zheng-ze(School of Mathematics and Physics,Anshun University,Anshun 561000,China)

机构地区：[1]安顺学院数理学院,贵州安顺561000

出　　处：《数学的实践与认识》2022年第3期81-88,共8页Mathematics in Practice and Theory

基　　金：贵州省教育厅自然科学研究资助项目(黔教合KY字[2018]337,[2018]034,[2018]070,[2019]069,[2020]131,[2020]136);贵州省科技厅科学技术基金课题(黔科合基础[2020]1Y278);国家自然科学基金(61762001)。

摘　　要：按照时间顺序,记粮食产量与影响它的生产条件为数据序列,根据灰色关联的基本原理计算他们之间的关联系数,通过对比系数值的大小,分别将和粮食产量关系密切的一些生产因素聚为一类,系数较小的一些因素聚为另一类.接着以两组数据为基础建立GM(1,n)预测模型,预测粮食的产量.发现第一组的预测值与实际统计值相对误差较小,且在模型精度规定范围之内,由此,说明可以通过灰色关联分析方法对影响粮食产量的多个生产因素降维、聚类、重组数据空间做研究.In chronological order,the grain yield and the production conditions that affect it are recorded as data series.According to the basic principle of Grey correlation,the correlation coefficient between them is calculated.By comparing the size of the coefficient values,Some production factors closely related to grain output were grouped into a group,Factors with smaller coefficients cluster into another category.Then the GM(1,n)forecasting model was established based on the two sets of data to predict the grain yield.It was found that the relative error between the predicted value and the actual statistical value of the first group was smaller,and within the specified range of the model accuracy.Therefore,it was indicated that the dimensional reduction,clustering and data space reorganization of multiple production factors affecting grain yield could be studied by using the Grey correlation analysis method.

关 键 词：灰关联 聚类 粮食产量 GM(1 N) 预测 

分 类 号：F326.11[经济管理—产业经济] TP311.13[自动化与计算机技术—计算机软件与理论]