Forest structure analysis is important for understanding the properties and development of a forest community,and its outcomes can be influenced by how trees are measured in sampled plots.Although there is a general c...Forest structure analysis is important for understanding the properties and development of a forest community,and its outcomes can be influenced by how trees are measured in sampled plots.Although there is a general consensus on the height at which tree diameter should be measured[1.3 m:diameter at breast height(DBH)],the minimum measureddiameter(MMD)often varies in different studies.In this study,we assumed that the outcomes of forest structure analysis can be influenced by MMD and,to this end,we applied g(r)function and stand spatial structural parameters(SSSPs)to investigate how different MMDs affect forest spatial structure analysis in two pine-oak mixed forests(30 and 57 years old)in southwest China and one old-growth oak forest(>120years old)from northwest China.Our results showed that 1)MMD was closely related to the distribution patterns of forest trees.Tree distribution patterns at each observational scale(r=0-20 m)tended tobecome random as the MMD increased.The older the community,the earlier this random distribution pattern appeared.2)As the MMD increased,neighboring trees became more regularly distributed around a reference tree.In most cases,however,nearest neighbors of a reference tree were randomly distributed.3)Tree species mingling decreased with increasing diameter,but it decreased slowly in older forests.4)No correlations can be found between individual tree size differentiation and MMD.We recommend that comparisons of spatial structures between communities would be more effective if using a unified MMD criterion.展开更多
累计局部离群因子(cumulative local outlier factor,C_LOF)算法能有效解决数据流中的概念漂移问题和克服离群点检测中的伪装问题,但在处理高维数据时,时间复杂度较高。为有效解决时间复杂度高的问题,提出一种基于投影索引近邻的累计局...累计局部离群因子(cumulative local outlier factor,C_LOF)算法能有效解决数据流中的概念漂移问题和克服离群点检测中的伪装问题,但在处理高维数据时,时间复杂度较高。为有效解决时间复杂度高的问题,提出一种基于投影索引近邻的累计局部离群因子(cumulative local outlier factor based projection indexed nearest neighbor,PINN_C_LOF)算法。使用滑动窗口维护活跃数据点,在新数据到达和旧数据过期时,引入投影索引近邻(projection indexed nearest neighbor,PINN)方法,增量更新窗口中受影响数据点的近邻。实验结果表明,PINN_C_LOF算法在检测高维流数据离群值时,在保持检测精确度的前提下,其时间复杂度较C_LOF算法明显降低。展开更多
基金financially supported by the National Science Foundation of China (grant no. 31400542 31460196)+1 种基金Guangxi Natural Science Foundation (grant 2016GXNSFBA380233)Guangxi special fund project for innovation-driven development (AA 17204087-8)
文摘Forest structure analysis is important for understanding the properties and development of a forest community,and its outcomes can be influenced by how trees are measured in sampled plots.Although there is a general consensus on the height at which tree diameter should be measured[1.3 m:diameter at breast height(DBH)],the minimum measureddiameter(MMD)often varies in different studies.In this study,we assumed that the outcomes of forest structure analysis can be influenced by MMD and,to this end,we applied g(r)function and stand spatial structural parameters(SSSPs)to investigate how different MMDs affect forest spatial structure analysis in two pine-oak mixed forests(30 and 57 years old)in southwest China and one old-growth oak forest(>120years old)from northwest China.Our results showed that 1)MMD was closely related to the distribution patterns of forest trees.Tree distribution patterns at each observational scale(r=0-20 m)tended tobecome random as the MMD increased.The older the community,the earlier this random distribution pattern appeared.2)As the MMD increased,neighboring trees became more regularly distributed around a reference tree.In most cases,however,nearest neighbors of a reference tree were randomly distributed.3)Tree species mingling decreased with increasing diameter,but it decreased slowly in older forests.4)No correlations can be found between individual tree size differentiation and MMD.We recommend that comparisons of spatial structures between communities would be more effective if using a unified MMD criterion.
文摘累计局部离群因子(cumulative local outlier factor,C_LOF)算法能有效解决数据流中的概念漂移问题和克服离群点检测中的伪装问题,但在处理高维数据时,时间复杂度较高。为有效解决时间复杂度高的问题,提出一种基于投影索引近邻的累计局部离群因子(cumulative local outlier factor based projection indexed nearest neighbor,PINN_C_LOF)算法。使用滑动窗口维护活跃数据点,在新数据到达和旧数据过期时,引入投影索引近邻(projection indexed nearest neighbor,PINN)方法,增量更新窗口中受影响数据点的近邻。实验结果表明,PINN_C_LOF算法在检测高维流数据离群值时,在保持检测精确度的前提下,其时间复杂度较C_LOF算法明显降低。
