Joint Probability Estimation Using Tensor Decomposition and Dictionaries

TL;DR

This paper introduces a novel semi-parametric method combining dictionary representations and tensor decompositions for estimating joint probability densities of mixed discrete and continuous variables, outperforming traditional models.

Contribution

It is the first to integrate dictionaries with tensor decompositions for joint PDF estimation, effectively handling hybrid distributions and diverse data families.

Findings

Achieves better classification accuracy than state-of-the-art methods.

Demonstrates lower error rates on synthetic and real datasets.

Effectively models hybrid discrete-continuous distributions.

Abstract

In this work, we study non-parametric estimation of joint probabilities of a given set of discrete and continuous random variables from their (empirically estimated) 2D marginals, under the assumption that the joint probability could be decomposed and approximated by a mixture of product densities/mass functions. The problem of estimating the joint probability density function (PDF) using semi-parametric techniques such as Gaussian Mixture Models (GMMs) is widely studied. However such techniques yield poor results when the underlying densities are mixtures of various other families of distributions such as Laplacian or generalized Gaussian, uniform, Cauchy, etc. Further, GMMs are not the best choice to estimate joint distributions which are hybrid in nature, i.e., some random variables are discrete while others are continuous. We present a novel approach for estimating the PDF using ideas from dictionary representations in signal processing coupled with low rank tensor decompositions. To the best our knowledge, this is the first work on estimating joint PDFs employing dictionaries alongside tensor decompositions. We create a dictionary of various families of distributions by inspecting the data, and use it to approximate each decomposed factor of the product in the mixture. Our approach can naturally handle hybrid $N$-dimensional distributions. We test our approach on a variety of synthetic and real datasets to demonstrate its effectiveness in terms of better classification rates and lower error rates, when compared to state of the art estimators.