You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The greedy set cover algorithm is well-studied and the following theorem is known.
218
219
\begin{theorem}
219
220
If there exists a fractional set cover for $S$, using $p$ centers, then $GREEDYSETCOVER(G,S)$ outputs a cover of size at most $pH(\frac{\vert S \vert}{p})$.
220
221
\end{theorem}
221
222
Finally we give our REDUCE algorithm as follows:\\
222
223
\begin{definition}
223
224
In a directed graph $G$, $v$ is a center capturing vertex (CCV) if $\Gamma^-(v) \subset\Gamma^+(v)$.
\section{The $k$-centers Problem under Perturbation Resilience}
239
242
240
243
Up to this point, we have discussed the symmetric and asymmetric $k$-centers problem with the object function defined as the maximum distance from the centers to the points. In this section, we introduce \emph{perturbation resilience} as introduced by Bilu and Linial and summarize two important results. First, Balcan and Liang's result for the symmetric problem under $(1+\sqrt2)$-perturbation resilience can be improved upon. Second, the symmetric problem under $(2-\epsilon)$-approximation stability is $NP$-hard.
0 commit comments