A low variance consistent test of relative dependency
arXiv:1406.3852
Abstract
We describe a novel non-parametric statistical hypothesis test of relative dependence between a source variable and two candidate target variables. Such a test enables us to determine whether one source variable is significantly more dependent on a first target variable or a second. Dependence is measured via the Hilbert-Schmidt Independence Criterion (HSIC), resulting in a pair of empirical dependence measures (source-target 1, source-target 2). We test whether the first dependence measure is significantly larger than the second. Modeling the covariance between these HSIC statistics leads to a provably more powerful test than the construction of independent HSIC statistics by sub-sampling. The resulting test is consistent and unbiased, and (being based on U-statistics) has favorable convergence properties. The test can be computed in quadratic time, matching the computational complexity of standard empirical HSIC estimators. The effectiveness of the test is demonstrated on several real-world problems: we identify language groups from a multilingual corpus, and we prove that tumor location is more dependent on gene expression than chromosomal imbalances. Source code is available for download at https://github.com/wbounliphone/reldep.
International Conference on Machine Learning, Jul 2015, Lille, France
References in corpus (8)
- Equitability, mutual information, and the maximal information coefficient
- Equivalence of distance-based and RKHS-based statistics in hypothesis testing
- Algorithms for Learning Kernels Based on Centered Alignment
- Kernel-based Conditional Independence Test and Application in Causal Discovery
- A consistent multivariate test of association based on ranks of distances
- A Kernel Test for Three-Variable Interactions
- A simpler condition for consistency of a kernel independence test
- A low variance consistent test of relative dependency