Fisher Exact Scanning for Dependency
From MaRDI portal
Publication:5229908
Abstract: We introduce a method---called Fisher exact scanning (FES)---for testing and identifying variable dependency that generalizes Fisher's exact test on contingency tables to contingency tables and continuous sample spaces. FES proceeds through scanning over the sample space using windows in the form of tables of various sizes, and on each window completing a Fisher's exact test. Based on a factorization of Fisher's multivariate hypergeometric (MHG) likelihood into the product of the univariate hypergeometric likelihoods, we show that there exists a coarse-to-fine, sequential generative representation for the MHG model in the form of a Bayesian network, which in turn implies the mutual independence (up to deviation due to discreteness) among the Fisher's exact tests completed under FES. This allows an exact characterization of the joint null distribution of the -values and gives rise to an effective inference recipe through simple multiple testing procedures such as v{S}id'{a}k and Bonferroni corrections, eliminating the need for resampling. In addition, FES can characterize dependency through reporting significant windows after multiple testing control. The computational complexity of FES is approximately linear in the sample size, which along with the avoidance of resampling makes it ideal for analyzing massive data sets. We use extensive numerical studies to illustrate the work of FES and compare it to several state-of-the-art methods for testing dependency in both statistical and computational performance. Finally, we apply FES to analyzing a microbiome data set and further investigate its relationship with other popular dependency metrics in that context.
Recommendations
- Multi-scale Fisher’s independence test for multivariate dependence
- New upper bounds for tight and fast approximation of Fisher's exact test in dependency rule mining
- A general framework for multiple testing dependence
- The Generalized Fisher's Combination and Accurate P-Value Calculation under Dependence
- A generalization of Fisher's exact test in p×q contingency tables using more concordant relations
Cites work
- scientific article; zbMATH DE number 6107964 (Why is no real title available?)
- A Bayesian nonparametric approach to testing for dependence between random variables
- A Non-Parametric Test of Independence
- A consistent multivariate test of association based on ranks of distances
- ALGORITHM 643
- Adaptive shrinkage in Pólya tree type models
- BET on independence
- Brownian distance covariance
- Consistent distribution-free \(K\)-sample and independence tests for univariate random variables
- Detecting novel associations in large data sets
- Equitability, mutual information, and the maximal information coefficient
- Generalized R-squared for detecting dependence
- Measuring and testing dependence by correlation of distances
- Multiscale likelihood analysis and complexity penalized estimation.
- Optimal and fast detection of spatial clusters with scan statistics
- THE COMBINATION OF PROBABILITIES ARISING FROM DATA IN DISCRETE DISTRIBUTIONS
Cited in
(9)- BET on independence
- New upper bounds for tight and fast approximation of Fisher's exact test in dependency rule mining
- The Binary Expansion Randomized Ensemble Test
- Nonparametric Scanning Tests of Homogeneity for Hierarchical Models with Continuous Covariates
- Comments on “A Gibbs Sampler for a Class of Random Convex Polytopes”
- A new framework for distance and kernel-based metrics in high dimensions
- Bayesian nonparametric test for independence between random vectors
- ODC and ROC curves, comparison curves and stochastic dominance
- Distribution-Free Consistent Independence Tests via Center-Outward Ranks and Signs
This page was built for publication: Fisher Exact Scanning for Dependency
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5229908)
