TAAFT
Free mode
100% free
Freemium
Free Trial
Create tool

UMAP

[juː mæp]
Machine Learning
Last updated: December 9, 2024

Definition

A dimensionality reduction algorithm that preserves both local and global data structure.

Detailed Explanation

Uniform Manifold Approximation and Projection is a manifold learning technique for dimension reduction based on Riemannian geometry and algebraic topology. It often provides better preservation of global structure than t-SNE while maintaining computational efficiency.

Use Cases

Single-cell RNA sequencing analysis, large-scale data visualization, clustering visualization

Related Terms