On UMAP's True Loss Function
–Neural Information Processing Systems
In this work, we investigate UMAP's sampling based optimization scheme in detail. We derive UMAP's true loss function in closed form and find that it differs from
Neural Information Processing Systems
Oct-3-2025, 04:42:16 GMT
- Country:
- Europe > Germany
- Baden-Württemberg > Karlsruhe Region > Heidelberg (0.04)
- North America > United States
- Massachusetts > Middlesex County > Cambridge (0.04)
- Europe > Germany
- Technology: