2019
On Euclidean k-Means Clustering with alpha-Center Proximity
AISTATS 2019poster
$k$-means clustering is NP-hard in the worst case but previous work has shown efficient algorithms assuming the optimal $k$-means clusters are \emph{stable} under additive or multiplicative perturbation of data. This has two caveats. First, we do not know how to efficiently verify this property of o…