A robust sparse clustering method uses spatial medians and automatic feature exclusion to achieve competitive accuracy and better stability than standard K-means on simulated heavy-tailed high-dimensional data.
IEEE Transactions on Pattern Analysis and Machine Intelligence , title=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Sparse $K$-spatial-median clustering for high-dimensional data
A robust sparse clustering method uses spatial medians and automatic feature exclusion to achieve competitive accuracy and better stability than standard K-means on simulated heavy-tailed high-dimensional data.