Visualizing probabilistic models in Minkowski space with intensive symmetrized Kullback-Leibler embedding

TL;DR

This paper introduces a method to embed and visualize exponential family probabilistic models in Minkowski space using symmetrized Kullback-Leibler divergence, aiding understanding of model behavior and complexity.

Contribution

It provides explicit formulas for isometric embeddings of exponential family models into Minkowski space, enabling visualization of model predictions and analysis of parameter sufficiency.

Findings

Embedded models preserve natural distances via symmetrized Kullback-Leibler divergence.

Visualizations include Bernoulli, ideal gas, die, and Gaussian models.

Application to Ising model reveals manifold behavior near critical points.

Abstract

We show that the predicted probability distributions for any $N$-parameter statistical model taking the form of an exponential family can be explicitly and analytically embedded isometrically in a $N{+}N$-dimensional Minkowski space. That is, the model predictions can be visualized as control parameters are varied, preserving the natural distance between probability distributions. All pairwise distances between model instances are given by the symmetrized Kullback-Leibler divergence. We give formulas for these intensive symmetrized Kullback Leibler (isKL) coordinate embeddings, and illustrate the resulting visualizations with the Bernoulli (coin toss) problem, the ideal gas, $n$ sided die, the nonlinear least squares fit, and the Gaussian fit. We highlight how isKL can be used to determine the minimum number of parameters needed to describe probabilistic data, and conclude by visualizing the prediction space of the two-dimensional Ising model, where we examine the manifold behavior near its critical point.