random projection to O(log n/ε²) dimensions preserves pairwise distances within 1±ε

Can you shrink a 3D object into 2D without losing its shape?

Imagine you're packing a suitcase for a trip, but you want to fit everything in a smaller bag. You don't want to leave out any important items or change how they fit together.

Think of random projection as a magic trick that squishes a 3D object into a 2D space while keeping its shape intact. It's like folding a piece of paper in a way that it still looks like the original shape when you unfold it.

Example

You have a suitcase with 3D objects like a ball, a cube, and a pyramid. Using random projection, you can fold them into a flat 2D paper version that looks like the original 3D shapes.

Remember this

Random projection helps you fit a 3D object into a 2D space without losing its original shape.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

the Johnson-Lindenstrauss lemma says

Can we shrink big data without losing important details?

List of unsolved problems in mathematics

Why do random points in high dimensions seem to be evenly spaced?

orthogonal matrices preserve distances: O^T O = I means no stretching or squashing

How can you stretch or squash a square without changing its shape?

cosine similarity works better than Euclidean distance in high dimensions

Cosine similarity measures orientation, not magnitude, making it more robust to irrelevant dimensions in high-dimensional spaces

t-SNE preserves local structure

Can we see the hidden patterns in a cloud of data points?

Ordinary least squares

OLS minimizes squared differences

Swipe through 100 ML concepts daily

Open Pocket Polymath