Property testing of joint distributions using conditional samples

DOI10.1145/3241377MaRDI QIDQ5205798zbMATH OpenOpenAlexFDO

Authors Rishiraj Bhattacharyya, Sourav Chakraborty

Publication date 16 December 2019

Published in ACM Transactions on Computation Theory (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1702.01454

zbMATH Keywords

conditional samplingdistribution testingsubcube conditioning

Mathematics Subject Classification ID

Probability in computer science (algorithm analysis, random structures, phase transitions, etc.) (68Q87) Randomized algorithms (68W20)

Abstract: In this paper, we consider the problem of testing properties of joint distributions under the Conditional Sampling framework. In the standard sampling model, the sample complexity of testing properties of joint distributions is exponential in the dimension, resulting in inefficient algorithms for practical use. While recent results achieve efficient algorithms for product distributions with significantly smaller sample complexity, no efficient algorithm is expected when the marginals are not independent. We initialize the study of conditional sampling in the multidimensional setting. We propose a subcube conditional sampling model where the tester can condition on an (adaptively) chosen subcube of the domain. Due to its simplicity, this model is potentially implementable in many practical applications, particularly when the distribution is a joint distribution over $Sigm{a}^n$ for some set $Sigma$. We present algorithms for various fundamental properties of distributions in the subcube-conditioning model and prove that the sample complexity is polynomial in the dimension $n$ (and not exponential as in the traditional model). We present an algorithm for testing identity to a known distribution using $ildemathcalO(n^2)$-subcube-conditional samples, an algorithm for testing identity between two unknown distributions using $ildemathcalO(n^5)$-subcube-conditional samples and an algorithm for testing identity to a product distribution using $tildemathcalO(n^5)$-subcube-conditional samples. The central concept of our technique involves an elegant chain rule which can be proved using basic techniques of probability theory yet powerful enough to avoid the curse of dimensionality.

Recommendations

• Testing probability distributions using conditional samples
• Testing equivalence between distributions using conditional samples
• Improving and extending the testing of distributions for shape-restricted properties

Cited in

(6)

• Efficient parameter estimation of truncated Boolean product distributions
• Property Testing on Product Distributions
• Testing conditional independence of discrete distributions
• Property Testing on Product Distributions: Optimal Testers for Bounded Derivative Properties
• Lifting uniform learners via distributional decomposition
• Complexity of high-dimensional identity testing with coordinate conditional sampling

This page was built for publication: Property testing of joint distributions using conditional samples

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5205798)