The Randomized Dependence Coefficient
About
We introduce the Randomized Dependence Coefficient (RDC), a measure of non-linear dependence between random variables of arbitrary dimension based on the Hirschfeld-Gebelein-R\'enyi Maximum Correlation Coefficient. RDC is defined in terms of correlation of random non-linear copula projections; it is invariant with respect to marginal distribution transformations, has low computational cost and is easy to implement: just five lines of R code, included at the end of the paper.
David Lopez-Paz, Philipp Hennig, Bernhard Sch\"olkopf• 2013
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Dependence Detection | Linear dependence model | Power28 | 25 | |
| Dependence Testing | Laplace A | Power23.1 | 22 | |
| Regression | Real-data benchmark Regression (8 datasets) | Main Rank8.83 | 19 | |
| Classification | 22 Real-World Classification Datasets | Mean Rank12.29 | 17 | |
| Dependence Testing | Tree ring B | Power0.466 | 2 |
Showing 5 of 5 rows