t-Distributed Stochastic Neighbour Embeddings

What is t-SNE?

t-Distributed Stochastic Neighbor Embedding (commonly known as t-SNE) is a dimension reduction technique that preserves relative distances between points in high-dimensional data.

Recommended Path for Learning

• General overview, given by one of the creators of t-SNE:

• An in-depth breakdown (with an example in Python!)

Further Learning

Parameter adjustments can be tricky! Here’s an interactive resources for How to Use t-SNE Effectively.

Video

• Simplified explanation (via StatQuest):

Applied papers

• Paper 1
• Paper 2

Online tutorials

• Here is a simple, yet comprehensive tutorial

• If you have less time, this very basic R tutorial can get you visualizations in under 5 minutes

Theory papers

• Paper 1
• Paper 2