Markov Blanket Ranking using Kernel-based Conditional Dependence Measures

@inproceedings{Strobl2019MarkovBR,
  title={Markov Blanket Ranking using Kernel-based Conditional Dependence Measures},
  author={Eric V. Strobl and S. Visweswaran},
  booktitle={Cause Effect Pairs in Machine Learning},
  year={2019}
}
  • Eric V. Strobl, S. Visweswaran
  • Published in
    Cause Effect Pairs in Machine…
    2019
  • Computer Science, Mathematics
  • Developing feature selection algorithms that move beyond a pure correlational to a more causal analysis of observational data is an important problem in the sciences. Several algorithms attempt to do so by discovering the Markov blanket of a target, but they all contain a forward selection step which variables must pass in order to be included in the conditioning set. As a result, these algorithms may not consider all possible conditional multivariate combinations. We improve on this limitation… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 23 REFERENCES
    Feature Selection by Approximating the Markov Blanket in a Kernel-Induced Space
    9
    Kernel-based Conditional Independence Test and Application in Causal Discovery
    275
    Algorithms for discovery of multiple Markov boundaries
    45
    A Linear Non-Gaussian Acyclic Model for Causal Discovery
    708
    HITON: A Novel Markov Blanket Algorithm for Optimal Variable Selection
    248
    An Anytime Algorithm for Causal Inference
    54
    Kernel dimension reduction in regression
    235
    Computational Methods of Feature Selection
    924
    The max-min hill-climbing Bayesian network structure learning algorithm
    1149