arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1902.08559cs.DS
收藏

参数化k-聚类:距离很重要!

Parameterized k-Clustering: The distance matters!

Fedor V. Fomin, Petr A. Golovach, Kirill Simonov

摘要:

我们认为$k$ -聚类问题,这是一个给定的多集的$n$向量$X\subset \mathbb{Z}^d$和一个非负数$D$,以决定是否$X$可以被划分成$k$集群$C_1, \dots, C_k$,使成本\[\sum_{i=1}^k \min_{c_i\in \mathbb{R}^d}\sum_{x \in C_i} \|x-c_i\|_p^p \leq D,\],其中$\|\cdot\|_p$是Minkowski($L_p$)规范的顺序$p$。对于$p=1$,$k$ -Clustering是众所周知的$k$ -Median。对于$p=2$,在欧几里德距离的情况下,$k$ -聚类是$k$ -均值。我们发现,$k$ -聚类的参数化复杂性强烈依赖于距离顺序$p$。特别是,我们证明,对于每一个$p\in (0,1]$,$k$ -聚类是可解的时间$2^{O(D \log{D})} (nd)^{O(1)}$,因此是固定参数易处理的$D$参数化时。另一方面,我们证明了对于$p=0$和$p=\infty$阶的距离,不存在这样的算法,除非FPT=W[1]。

英文摘要:

We consider the $k$-Clustering problem, which is for a given multiset of $n$ vectors $X\subset \mathbb{Z}^d$ and a nonnegative number $D$, to decide whether $X$ can be partitioned into $k$ clusters $C_1, \dots, C_k$ such that the cost \[\sum_{i=1}^k \min_{c_i\in \mathbb{R}^d}\sum_{x \in C_i} \|x-c_i\|_p^p \leq D,\] where $\|\cdot\|_p$ is the Minkowski ($L_p$) norm of order $p$. For $p=1$, $k$-Clustering is the well-known $k$-Median. For $p=2$, the case of the Euclidean distance, $k$-Clustering is $k$-Means. We show that the parameterized complexity of $k$-Clustering strongly depends on the distance order $p$. In particular, we prove that for every $p\in (0,1]$, $k$-Clustering is solvable in time $2^{O(D \log{D})} (nd)^{O(1)}$, and hence is fixed-parameter tractable when parameterized by $D$. On the other hand, we prove that for distances of orders $p=0$ and $p=\infty$, no such algorithm exists, unless FPT=W[1].

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。

↑