dimensionality reduction

Definitions

1. A set of techniques (PCA, t-SNE, UMAP) that reduce the number of features in a dataset while preserving as much meaningful structure as possible. Essential when dealing with high-dimensional data where visualization is impossible and models suffer from the curse of dimensionality. Also used to compress embeddings and speed up training.

   In plain English: Squishing data with hundreds of measurements down to just a few key dimensions so you can visualize it or help a model learn faster without drowning in too many numbers.

   Example: They used UMAP to project the 768-dimensional embeddings down to 2D, revealing clear clusters of similar documents in the scatter plot.

Related Terms

• Embedding
• regression
• clustering