检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:刘晓晓[1,2] 叶持跃[1,2] 李加林[1,2] 徐亚东[1] 庄汝龙 宓科娜
机构地区:[1]宁波大学城市科学系,中国浙江宁波315211 [2]宁波大学浙江省海洋文化与经济研究中心,中国浙江宁波315211
出 处:《经济地理》2014年第6期87-91,共5页Economic Geography
基 金:“十二五”国家科技支撑计划项目(2013BAJ10B06-01);宁波市农业择优委托项目(2011C11008);宁波大学科研基金(理)/学科项目(xkl12013)
摘 要:周-布方法是影响较大的城市职能分类方法,SOFM网络方法是应用人工神经网络进行城市职能分类的新方法。采用这两种方法分别将2010年135个城市的职能分为3类、5类和12类。比较二者的异同点和优缺点,得出:①周-布方法是监督聚类,权重设置对分类结果有很大影响,SOFM网络方法是非监督聚类,避免了人为确定指标的主观性;②周-布方法计算速率快,分类结果稳定,分类类别数较易确定;③周-布方法的三种分类结果之间形成聚类树形图,SOFM网络方法的三种结果之间存在交叉;④周-布方法对规模数据比较敏感,SOFM网络方法对行业数据比较敏感。⑤周-布方法的多个城市分类不合理,SOFM网络方法职能类型较明显,分类结果更为合理。综合来看,SOFM网络方法优于周-布方法,但分类结果的稳定性会影响其使用范围,需进一步研究。Zhou Yixing-R. Bradshaw Method(Zhou-Bradshaw Method) has the greatest effect in urban functional classification, while Self-Organizing Feature Map Network Method(SOFM Network Method) is a new method by applying artificial neural networks(ANN). In this paper, 135 cities, in 2010, were divided into 3, 5, 12 classes separately with the two methods. Comparing the similarities and differences between two methods and the advantages and disadvantages, it concluded that: (1)Zhou-Bradshaw Method is a supervised clustering, and weight setting had great impact on the classification results. SOFM Network Method is an unsupervised clustering, and it avoided determining the index weights. (2)The computation rate of Zhou-Bradshaw Method was faster, the classification results are stable, and the numbers of classification categories are were easier to be determined. (3)Three kinds of classification results of Zhou- Bradshaw Method form a clustering tree diagram, while the results of SOFM Network Method were crossed. (4)Zhou- Bradshaw Method was sensitive about the date of scale index, SOFM Network Method was sensitive about the date of industry index. (5)The results of cities with Zhou-Bradshaw Method were unreasonable. The urban fimctional characteristics of SOFM Network Method were more obvious and the results were more reasonable.Comprehensively, SOFM Network Method is better than Zhou-Bradshaw Method. But the stability of results will affect the using range of the method, so it need further study.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:18.116.87.126