FFML, TRFF, and FFCI are practical RFF-based approximations that replace expensive GP kernel matrices with finite feature maps, delivering competitive precision-recall trade-offs for score-based and constraint-based causal discovery in nonlinear mixed data.
Advances in Neural Information Processing Systems , volume=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Fourier Feature Methods for Nonlinear Causal Discovery: FFML Scoring, TRFF Scoring, and FFCI Testing in Mixed Data
FFML, TRFF, and FFCI are practical RFF-based approximations that replace expensive GP kernel matrices with finite feature maps, delivering competitive precision-recall trade-offs for score-based and constraint-based causal discovery in nonlinear mixed data.