UMAP
Cross-source consensus on UMAP from 1 sources and 5 claims.
1 sources · 5 claims
Risks & contraindications
Comparisons
Evidence quality
Highlighted claims
- cuML UMAP had topology error 5 on all 11 OpenML datasets in the benchmark. — DiRe-RAPIDS: Topology-faithful dimensionality reduction at scale
- UMAP scored higher than DiRe on local-neighborhood metrics at k = 16 in the arXiv experiment. — DiRe-RAPIDS: Topology-faithful dimensionality reduction at scale
- UMAP partially tracked noise-induced beta1 inflation in the noisy figure-8 experiment. — DiRe-RAPIDS: Topology-faithful dimensionality reduction at scale
- The article cautions against treating UMAP islands and gaps as direct evidence of real scientific separations. — DiRe-RAPIDS: Topology-faithful dimensionality reduction at scale
- UMAP’s local-neighborhood objective can conflict with global and topological structure. — DiRe-RAPIDS: Topology-faithful dimensionality reduction at scale