The overall NDVI characteristics and precipitation sensitivity in the study area from 2000 to 2018 were investigated using NDVI images of urban agglomeration in central Shanxi basin in 2000,2005,2010,and 2018 as well ...The overall NDVI characteristics and precipitation sensitivity in the study area from 2000 to 2018 were investigated using NDVI images of urban agglomeration in central Shanxi basin in 2000,2005,2010,and 2018 as well as the climate data of China’s surface cumulative annual value data set(1981-2010)in accordance with the method such as the geographically weighted regression model.As can be seen from the results,first,the overall NDVI pattern of urban agglomeration in central Shanxi basin,China has been changed and distributed along the topography in the shape of strip.Second,the spatial evolution of NDVI in the study area is varied significantly with the trend of expansion as a whole and a shrinking trend in some areas.Third,the overall precipitation in the study area presents a declining trend from the west to the east,while the NDVI precipitation sensitivity shows a decreasing trend from west to east.To be specific,the west and the east have a positive value of 1.3129 with strong sensitivity and a negative value of-1.0908 with weak sensitivity,respectively.The study results are expected to provide a scientific basis for restoring vegetation and formulating disaster prevention policies for urban agglomeration in Central Shanxi Basin.展开更多
长期以来,孤立点的检测一直聚焦于簇边缘的离散点,当聚类后簇的数目低于实际数目,或孤立点被伪装在簇内的情况下,簇内孤立点的判定则会更加困难.为判定簇内孤立点,提出一种基于密度聚类DBSCAN (Density based spatial clustering of app...长期以来,孤立点的检测一直聚焦于簇边缘的离散点,当聚类后簇的数目低于实际数目,或孤立点被伪装在簇内的情况下,簇内孤立点的判定则会更加困难.为判定簇内孤立点,提出一种基于密度聚类DBSCAN (Density based spatial clustering of application with noise)的簇内孤立点检测方法 ODIC-DBSCAN (Outlier detection of inner-cluster based on DBSCAN).首先在建立距离矩阵的基础上,通过半径获取策略得到针对该点集的k个有效半径Radius集合,并据此构造密度矩阵;然后建立点集覆盖模型,提出了相邻有效半径构造的覆盖多维体能够覆盖点集的思想,并通过拉格朗日乘子法求取最优的覆盖多维体数目之比,输出点比阈值组;最后重建ODIC-DBSCAN的孤立点检测方法,以簇发生融合现象作为算法终止的判定条件.实验通过模拟数据集,公开benchmark与UCI数据集共同验证了ODIC-DBSCAN算法,展示了聚类过程;分析了算法性能;并通过与其他聚类、孤立点判定方法的对比,验证了算法对簇内孤立点的判定效果.展开更多
基金Project(cstc2021ycjh-bgzxm0165) supported by the Natural Science Foundation of Chongqing,ChinaProject(BWLCF_(2)02102) supported by the China Baowu Low Carbon Metallurgy Innovation FoundationProject(CYB20007) supported by the Graduate Scientific Research and Innovation Foundation of Chongqing,China。
基金supported by the National Natural Science Foundation of China(Grant No.41701223)。
文摘The overall NDVI characteristics and precipitation sensitivity in the study area from 2000 to 2018 were investigated using NDVI images of urban agglomeration in central Shanxi basin in 2000,2005,2010,and 2018 as well as the climate data of China’s surface cumulative annual value data set(1981-2010)in accordance with the method such as the geographically weighted regression model.As can be seen from the results,first,the overall NDVI pattern of urban agglomeration in central Shanxi basin,China has been changed and distributed along the topography in the shape of strip.Second,the spatial evolution of NDVI in the study area is varied significantly with the trend of expansion as a whole and a shrinking trend in some areas.Third,the overall precipitation in the study area presents a declining trend from the west to the east,while the NDVI precipitation sensitivity shows a decreasing trend from west to east.To be specific,the west and the east have a positive value of 1.3129 with strong sensitivity and a negative value of-1.0908 with weak sensitivity,respectively.The study results are expected to provide a scientific basis for restoring vegetation and formulating disaster prevention policies for urban agglomeration in Central Shanxi Basin.
文摘长期以来,孤立点的检测一直聚焦于簇边缘的离散点,当聚类后簇的数目低于实际数目,或孤立点被伪装在簇内的情况下,簇内孤立点的判定则会更加困难.为判定簇内孤立点,提出一种基于密度聚类DBSCAN (Density based spatial clustering of application with noise)的簇内孤立点检测方法 ODIC-DBSCAN (Outlier detection of inner-cluster based on DBSCAN).首先在建立距离矩阵的基础上,通过半径获取策略得到针对该点集的k个有效半径Radius集合,并据此构造密度矩阵;然后建立点集覆盖模型,提出了相邻有效半径构造的覆盖多维体能够覆盖点集的思想,并通过拉格朗日乘子法求取最优的覆盖多维体数目之比,输出点比阈值组;最后重建ODIC-DBSCAN的孤立点检测方法,以簇发生融合现象作为算法终止的判定条件.实验通过模拟数据集,公开benchmark与UCI数据集共同验证了ODIC-DBSCAN算法,展示了聚类过程;分析了算法性能;并通过与其他聚类、孤立点判定方法的对比,验证了算法对簇内孤立点的判定效果.
基金Supported by National Natural Science Foundation of China(61874037,61505033)National Postdoctoral Science Foundation of China(2017M621254,2018T110280)+2 种基金Heilongjiang Provincial Postdoctoral Science Foundation(LBH-TZ1708)Key Laboratory of Micro-systems and Micro-structures Manufacturing of Ministry of Education,Harbin Institute of Technology(2017KM003)Fundamental Research Funds for The Central Universities(HIT.NSRIF.2019060)~~