Support-weighted partial recentering of maxmin seeds using halfspace depth yields consistent geometric improvement over standard maxmin in planar benchmarks while preserving thresholded H1 summaries.
Topological data analysis with ϵ-net induced lazy witness complexes
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Local Depth-Based Corrections to Maxmin Landmark Selection for Lazy Witness Persistence
Support-weighted partial recentering of maxmin seeds using halfspace depth yields consistent geometric improvement over standard maxmin in planar benchmarks while preserving thresholded H1 summaries.