A polynomial-time algorithm for learning nonparametric causal graphs
From MaRDI portal
Publication:6343409
arXiv2006.11970MaRDI QIDQ6343409
Author name not available (Why is that?)
Publication date: 21 June 2020
Abstract: We establish finite-sample guarantees for a polynomial-time algorithm for learning a nonlinear, nonparametric directed acyclic graphical (DAG) model from data. The analysis is model-free and does not assume linearity, additivity, independent noise, or faithfulness. Instead, we impose a condition on the residual variances that is closely related to previous work on linear models with equal variances. Compared to an optimal algorithm with oracle knowledge of the variable ordering, the additional cost of the algorithm is linear in the dimension and the number of samples . Finally, we compare the proposed algorithm to existing approaches in a simulation study.
Has companion code repository: https://github.com/MingGao97/NPVAR
This page was built for publication: A polynomial-time algorithm for learning nonparametric causal graphs
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6343409)