A two-stage manifold optimization method on the sphere uses Riemannian EM with a heavy-tailed kernel and projected density initialization to fit an unknown number of hyperplanes, claiming better geometric accuracy than prior baselines.
org/CorpusID:205939
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Fitting Unknown Number of Hyperplanes with Manifold Optimization
A two-stage manifold optimization method on the sphere uses Riemannian EM with a heavy-tailed kernel and projected density initialization to fit an unknown number of hyperplanes, claiming better geometric accuracy than prior baselines.