Machine Learning
Intermediate
4.5
Visualize High-Dimensional Data with t-SNE & UMAP
Turn embeddings into a 2D map you can actually see.
0h 20m
1 lesson
1.2K students
What You'll Learn
Learning objectives will be added soon.
Tutorial Content
Seeing structure
Embeddings live in hundreds of dimensions. t-SNE and UMAP project them to 2D so clusters and outliers become visible.
import umap
coords = umap.UMAP(n_neighbors=15, min_dist=0.1).fit_transform(vectors)Read them carefully
These plots preserve local structure well but can distort global distances and cluster sizes. Use them to explore and form hypotheses — not to measure exact distances. UMAP is usually faster and preserves more global structure than t-SNE.
Your Progress
Sign in to track your progress
Tags
Data Science
Embeddings
Machine Learning