arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1907.05094cs.DScs.CGcs.LG
收藏

Ward方法分析

Analysis of Ward's Method

University of Bonn(波恩大学)

Anna Großwendt, Heiko Röglin, Melanie Schmidt

摘要:

研究了求解分层$k$ -均值问题的Ward方法.这种流行的贪婪启发式算法基于\emph{complete linkage}范式:从所有数据点作为单例聚类开始,它连续合并两个聚类以形成少一个聚类的聚类。选择聚类对以(局部)最小化下一步聚类的$k$均值成本。完全连接算法是解决层次聚类问题的一种常用算法,但其理论性质的研究相对较少。对于欧几里德$k$ -中心问题,Ackermann等人证明了通过完全连接计算的层次中的$k$ -聚类具有$Θ(\log k)$的最坏情况近似比。如果数据位于$\mathbb{R}^d$中,则对于恒定维度$d$,保证提高到$\mathcal{O}(1)$,但是$\mathcal{O}$表示法隐藏了对$d$的线性依赖性。迄今为止,尚未分析$k$ -中位数或$k$ -均值的完全连锁。在本文中,我们表明,沃德的方法计算$2$ -近似相对于$k$ -意味着目标函数,如果最佳的$k$ -聚类分离良好。如果另外的最优聚类也满足平衡条件,则沃德的方法完全恢复最优解。这些结果在任意维数下都成立。我们伴随着我们的积极的结果与下限的$Ω((3/2)^d)$的数据集在$\mathbb{R}^d$,如果没有分离是保证,并与下限时,保证分离不够强。最后,我们表明,沃德产生一个$\mathcal{O}(1)$ -近似聚类的一维数据集。

英文摘要:

We study Ward's method for the hierarchical $k$-means problem. This popular greedy heuristic is based on the \emph{complete linkage} paradigm: Starting with all data points as singleton clusters, it successively merges two clusters to form a clustering with one cluster less. The pair of clusters is chosen to (locally) minimize the $k$-means cost of the clustering in the next step. Complete linkage algorithms are very popular for hierarchical clustering problems, yet their theoretical properties have been studied relatively little. For the Euclidean $k$-center problem, Ackermann et al. show that the $k$-clustering in the hierarchy computed by complete linkage has a worst-case approximation ratio of $Θ(\log k)$. If the data lies in $\mathbb{R}^d$ for constant dimension $d$, the guarantee improves to $\mathcal{O}(1)$, but the $\mathcal{O}$-notation hides a linear dependence on $d$. Complete linkage for $k$-median or $k$-means has not been analyzed so far. In this paper, we show that Ward's method computes a $2$-approximation with respect to the $k$-means objective function if the optimal $k$-clustering is well separated. If additionally the optimal clustering also satisfies a balance condition, then Ward's method fully recovers the optimum solution. These results hold in arbitrary dimension. We accompany our positive results with a lower bound of $Ω((3/2)^d)$ for data sets in $\mathbb{R}^d$ that holds if no separation is guaranteed, and with lower bounds when the guaranteed separation is not sufficiently strong. Finally, we show that Ward produces an $\mathcal{O}(1)$-approximative clustering for one-dimensional data sets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑