t-SNE, Forceful Colorings and Mean Field Limits

02/25/2021
by   Yulan Zhang, et al.
0

t-SNE is one of the most commonly used force-based nonlinear dimensionality reduction methods. This paper has two contributions: the first is forceful colorings, an idea that is also applicable to other force-based methods (UMAP, ForceAtlas2,...). In every equilibrium, the attractive and repulsive forces acting on a particle cancel out: however, both the size and the direction of the attractive (or repulsive) forces acting on a particle are related to its properties: the force vector can serve as an additional feature. Secondly, we analyze the case of t-SNE acting on a single homogeneous cluster (modeled by affinities coming from the adjacency matrix of a random k-regular graph); we derive a mean-field model that leads to interesting questions in classical calculus of variations. The model predicts that, in the limit, the t-SNE embedding of a single perfectly homogeneous cluster is not a point but a thin annulus of diameter ∼ k^-1/4 n^-1/4. This is supported by numerical results. The mean field ansatz extends to other force-based dimensionality reduction methods.

READ FULL TEXT

page 1

page 4

page 8

research
08/13/2022

May the force be with you

Modern methods in dimensionality reduction are dominated by nonlinear at...
research
07/09/2020

On the role of solute drag in reconciling laboratory and natural constraints on olivine grain growth kinetics

We investigate the effect of solute drag on grain growth (GG) kinetics i...
research
09/21/2020

Mean-field optimal control for biological pattern formation

We propose a mean-field optimal control problem for the parameter identi...
research
05/16/2022

Mean-Field Nonparametric Estimation of Interacting Particle Systems

This paper concerns the nonparametric estimation problem of the distribu...
research
06/02/2014

The constitution of visual perceptual units in the functional architecture of V1

Scope of this paper is to consider a mean field neural model which takes...
research
06/09/2020

Rank Reduction, Matrix Balancing, and Mean-Field Approximation on Statistical Manifold

We present a unified view of three different problems; rank reduction of...
research
07/17/2020

A Unifying Perspective on Neighbor Embeddings along the Attraction-Repulsion Spectrum

Neighbor embeddings are a family of methods for visualizing complex high...

Please sign up or login with your details

Forgot password? Click here to reset