Abstract:In this paper we introduce a new measure of conditional dependence between two random vectors ${\boldsymbol X}$ and ${\boldsymbol Y}$ given another random vector $\boldsymbol Z$ using the ball divergence. Our measure characterizes conditional independence and does not require any moment assumptions. We propose a consistent estimator of the measure using a kernel averaging technique and derive its asymptotic distribution. Using this statistic we construct two tests for conditional independence, one in the model-${\boldsymbol X}$ framework and the other based on a novel local wild bootstrap algorithm. In the model-${\boldsymbol X}$ framework, which assumes the knowledge of the distribution of ${\boldsymbol X}|{\boldsymbol Z}$, applying the conditional randomization test we obtain a method that controls Type I error in finite samples and is asymptotically consistent, even if the distribution of ${\boldsymbol X}|{\boldsymbol Z}$ is incorrectly specified up to distance preserving transformations. More generally, in situations where ${\boldsymbol X}|{\boldsymbol Z}$ is unknown or hard to estimate, we design a double-bandwidth based local wild bootstrap algorithm that asymptotically controls both Type I error and power. We illustrate the advantage of our method, both in terms of Type I error and power, in a range of simulation settings and also in a real data example. A consequence of our theoretical results is a general framework for studying the asymptotic properties of a 2-sample conditional $V$-statistic, which is of independent interest.

Analyzing Divergence for Nondeterministic Probabilistic Models

Introducing Divergence for Infinite Probabilistic Models

On Probabilistic Weak Bisimilarity

Analysis for Intransitive Noninterference Security Properties in Probabilistic Systems

Computing Marginal and Conditional Divergences between Decomposable Models with Applications

A Local Algorithm for Checking Probabilistic Bisimilarity

Probabilistic Bisimilarity As Testing Equivalence

Noninterference Analysis of Reversible Systems: An Approach Based on Branching Bisimilarity

A New Divergence Measure for Basic Probability Assignment and Its Applications in Extremely Uncertain Environments.

Probabilistic Weak Bisimulation and Axiomatization for Probabilistic Models.

Divergence Measure of Belief Function and Its Application in Data Fusion

R-divergence for Estimating Model-oriented Distribution Discrepancy

Information divergences of Markov chains and their applications

The Logarithmic Super Divergence and its use in Statistical Inference

A Ball Divergence Based Measure For Conditional Independence Testing

Similarity Query Processing for Probabilistic Sets

Bayesian estimation of the Kullback-Leibler divergence for categorical sytems using mixtures of Dirichlet priors

co-BPM: a Bayesian Model for Divergence Estimation

Amplifying Inter-Message Distance: On Information Divergence Measures in Big Data.

When Equivalence and Bisimulation Join Forces in Probabilistic Automata

Statistical and Topological Properties of Sliced Probability Divergences