An Analysis of the t-SNE Algorithm for Data Visualization

TL;DR

This paper provides the first formal analysis and provable guarantees for the effectiveness of t-SNE in producing meaningful 2D visualizations of clusterable data, under certain conditions.

Contribution

It introduces a formal framework for data visualization, offers rigorous analysis of t-SNE's performance under deterministic and probabilistic cluster conditions, and extends understanding of its capabilities.

Findings

t-SNE successfully separates clusters under certain deterministic conditions.

Probabilistic models like mixtures of log-concave distributions satisfy these conditions.

t-SNE can partially recover cluster structure even without the deterministic assumptions.

Abstract

A first line of attack in exploratory data analysis is data visualization, i.e., generating a 2-dimensional representation of data that makes clusters of similar points visually identifiable. Standard Johnson-Lindenstrauss dimensionality reduction does not produce data visualizations. The t-SNE heuristic of van der Maaten and Hinton, which is based on non-convex optimization, has become the de facto standard for visualization in a wide range of applications. This work gives a formal framework for the problem of data visualization - finding a 2-dimensional embedding of clusterable data that correctly separates individual clusters to make them visually identifiable. We then give a rigorous analysis of the performance of t-SNE under a natural, deterministic condition on the "ground-truth" clusters (similar to conditions assumed in earlier analyses of clustering) in the underlying data. These are the first provable guarantees on t-SNE for constructing good data visualizations. We show that our deterministic condition is satisfied by considerably general probabilistic generative models for clusterable data such as mixtures of well-separated log-concave distributions. Finally, we give theoretical evidence that t-SNE provably succeeds in partially recovering cluster structure even when the above deterministic condition is not met.