Stochastic Neighbor Embedding separates well-separated clusters

Abstract

Stochastic Neighbor Embedding and its variants are widely used dimensionality reduction techniques -- despite their popularity, no theoretical results are known. We prove that the optimal SNE embedding of well-separated clusters from high dimensions to any Euclidean space R^d manages to successfully separate the clusters in a quantitative way. The result also applies to a larger family of methods including a variant of t-SNE.

References in corpus (1)

• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems

Cited by in corpus (6)

• Theoretical Foundations of t-SNE for Visualizing High-Dimensional Clustered Data
• An Analysis of the t-SNE Algorithm for Data Visualization
• Manifold Learning in Atomistic Simulations: A Conceptual Review
• Modeling User Selection in Quality Diversity
• Improving the Effectiveness and Efficiency of Stochastic Neighbour Embedding with Isolation Kernel
• State-Driven Dynamic Graphon Model