UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
This provides a scalable and general-purpose dimension reduction method for machine learning applications, addressing limitations of existing techniques.
The paper tackles the problem of dimension reduction by introducing UMAP, a novel manifold learning technique that is competitive with t-SNE for visualization quality, preserves more global structure, and offers superior run time performance.
UMAP (Uniform Manifold Approximation and Projection) is a novel manifold learning technique for dimension reduction. UMAP is constructed from a theoretical framework based in Riemannian geometry and algebraic topology. The result is a practical scalable algorithm that applies to real world data. The UMAP algorithm is competitive with t-SNE for visualization quality, and arguably preserves more of the global structure with superior run time performance. Furthermore, UMAP has no computational restrictions on embedding dimension, making it viable as a general purpose dimension reduction technique for machine learning.