Variance misspecification in Gaussian mixtures creates a phase diagram: correct specification recovers true means independent of SNR, under-smoothing biases means with SNR^{-1} error in low SNR, and over-smoothing collapses clusters above an SNR-dependent threshold.
The catastrophic failure of the k-means algorithm in high dimen- sions, and how hartigan’s algorithm avoids it
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
A minor variant of Hartigan's k-means algorithm delivers 2-5% better clustering performance than the original, with gains increasing in high dimensions or large k.
citing papers explorer
-
The interplay of signal-to-noise ratio and variance misspecification in Gaussian mixtures
Variance misspecification in Gaussian mixtures creates a phase diagram: correct specification recovers true means independent of SNR, under-smoothing biases means with SNR^{-1} error in low SNR, and over-smoothing collapses clusters above an SNR-dependent threshold.
-
An effective variant of the Hartigan $k$-means algorithm
A minor variant of Hartigan's k-means algorithm delivers 2-5% better clustering performance than the original, with gains increasing in high dimensions or large k.