Machine Learning

UMAP

A fast dimensionality-reduction method that preserves local and global structure.

UMAP constructs a graph describing how observations are connected in high-dimensional space and then creates a lower-dimensional representation with a similar structure.

It is commonly used to visualise embeddings and large datasets. UMAP is often faster than t-SNE and may preserve more global structure, although its output is still affected by parameter choices.

Also in Machine Learning

Supervised Learning Unsupervised Learning K-Means K-Nearest Neighbours Logistic Regression Linear Regression

Apply

Begin the first module

Become AI native, it is the real deal today, and if it is not for you, you have lost nothing but learnt a new skill.

Apply