UMAP
UMAP, or Uniform Manifold Approximation and Projection, is a nonlinear manifold-learning technique that embeds high-dimensional data in a lower-dimensional space while preserving important neighborhood and structural relationships.
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
UMAP is a nonlinear dimensionality-reduction technique based on manifold-learning principles that constructs a low-dimensional representation of high-dimensional data.
- How it is used
- Supports exploratory data analysis by projecting complex multivariate data into a few dimensions for visualization.
- How deep it goes
- Defines nonlinear projection, applications, and caveats, but does not explain UMAP’s algorithm or implementation.
- Examples
- Uses low-dimensional charts to inspect patterns, clusters, and isolated points that may indicate anomalies.
- What you can do after
- Evaluate/understand complex datasets by inspecting exploratory projections for visible patterns and clusters.
Read how Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI explains UMAP →
Data Without Labels
Uniform Manifold Approximation and Projection (UMAP) is a nonlinear manifold-learning technique for dimensionality reduction that constructs a low-dimensional embedding while preserving data relationships.
- How it is used
- Deploys nonlinear projection for visualizing high-dimensional data and preprocessing datasets before modeling.
- How deep it goes
- Explains manifold assumptions, graph construction, cross-entropy optimization, stochastic gradient descent, and two principal parameters.
- Examples
- Uses gene-expression data, language embeddings, and image datasets to illustrate visualization and similarity-based grouping.
- What you can do after
- Apply it via umap-learn to visualize data, reduce dimensions, and support clustering or anomaly-detection analysis.
Graph Neural Networks in Action
UMAP is a nonlinear dimensionality-reduction technique that projects high-dimensional data into a lower-dimensional space while attempting to preserve local and global structure.
- How it is used
- Uses UMAP to visualize high-dimensional Node2Vec graph embeddings and inspect clusters, patterns, and relationships.
- How deep it goes
- Explains projection goals and tuning parameters, then demonstrates Python usage; stops short of mathematical derivation.
- Examples
- Visualizes Political Books Node2Vec embeddings and compares UMAP with direct two-dimensional Node2Vec visualization and t-SNE.
- What you can do after
- Apply it via Python to project embeddings into two dimensions and plot interpretable graph-structure visualizations.
Related concepts
- dimensionality reductionkind of · Dimensionality reduction is the transformation of data into a representation with fewer dimensions while retaining selected information, structure, or relationships.
- principal component analysisalternative to · Principal component analysis is a linear dimensionality-reduction technique that transforms data into orthogonal components ordered by the variance they capture.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.