Every point in high-dimensional space was grouped into some clusters using k-means cluster algorithm,then the weightee centroid distance of every point was computed based on the start distance and centroid distance of every point.
通过建模得到该DDM的四元组数据结构,对于高维空间中的数据点,通过k平均聚类算法将数据点聚成若干类,分别计算每个点对应的始点和质心距离,得到基于加权的质心距离,并将加权的质心距离作为每个数据点的索引键值,且用基于分片的B+树建立索引,得到了该索引的创建算法。
Two steps are made in HD-Tree, first for every character in high-dimensional space are grouped into T clusters using hierarchy-based cluster algorithm, then the uniform start distance and centroid distance of every characters are pre-calculated and indexed by a partition-based B+-tree.
首先将n个书法字通过层次聚类聚成若干类,然后分别计算每个字对应的统一化始点距离和质心距离,最后将两者结合生成索引键值。
We used centroid distance increment matrix to extract the CT heart shape characteristics.
本文利用质心距离增量矩阵方法对CT心脏图像的区域和边缘进行统一描述,再用质心距离增量矩阵方法提取CT心脏形状特征。
With several examples of single crystal structures of compounds incorporating aromatic rings,the criterion of the centroids distance of 0.
通过一些晶体结构的实例指出,目前广泛采用的质心距小于0。
CopyRight © 2020-2024 优校网[www.youxiaow.com]版权所有 All Rights Reserved. ICP备案号:浙ICP备2024058711号