Dependence Measure for non-additive model
arXiv:1310.1562
Abstract
We proposed a new statistical dependency measure called Copula Dependency Coefficient(CDC) for two sets of variables based on copula. It is robust to outliers, easy to implement, powerful and appropriate to high-dimensional variables. These properties are important in many applications. Experimental results show that CDC can detect the dependence between variables in both additive and non-additive models.
arXiv commentThis paper has been withdrawn by the author due to change of the main content
References in corpus (5)
- Brownian distance covariance
- A consistent multivariate test of association based on ranks of distances
- Comment on "Detecting Novel Associations In Large Data Sets" by Reshef Et Al, Science Dec 16, 2011
- Copula-based Kernel Dependency Measures
- A Simple Method for Obtaining the Maximal Correlation Coefficient and Related Characterizations