Conditional independence in max-linear Bayesian networks

@article{Amendola2022ConditionalII,
  title={Conditional independence in max-linear Bayesian networks},
  author={Carlos Am'endola and Claudia Kluppelberg and Steffen L. Lauritzen and Ngoc Mai Tran},
  journal={The Annals of Applied Probability},
  year={2022}
}
Motivated by extreme value theory, max-linear Bayesian networks have been recently introduced and studied as an alternative to linear structural equation models. However, for max-linear systems the classical independence results for Bayesian networks are far from exhausting valid conditional independence statements. We use tropical linear algebra to derive a compact representation of the conditional distribution given a partial observation, and exploit this to obtain a complete description of… 
Markov equivalence of max-linear Bayesian networks
TLDR
The parallel between the two theories via tropicalization is established, and the surprising result that the Markov equivalence classes for max-linear Bayesian networks coincide with the ones obtained by regular CI is established.
Recursive max-linear models with propagating noise
TLDR
A probabilistic analysis of this new noisy model reveals that the unique minimum DAG representing the distribution of the non-noisy model remains unchanged and identifiable, and it is proved that the estimated Kleene star matrix converges to a matrix of independent Weibull entries after proper centering and scaling.
Extremes of Markov random fields on block graphs
We study the joint occurrence of large values of a Markov random field or undirected graphical model associated to a block graph. On such graphs, containing trees as special cases, we aim to
Discrete max-linear Bayesian networks
TLDR
The fact that CBN models are toric varieties after linear change of coordinates to all discrete max-linear models is extended and similar results can be obtained.
Causal Discovery of a River Network from its Extremes
TLDR
QTree is provided, a new and simple algorithm to solve the Hidden River Problem that outperforms existing methods and relies on qualitative aspects of the max-linear Bayesian network model.
Estimating a Latent Tree for Extremes
TLDR
QTree is given, a simple and efficient algorithm to solve the Latent River Problem that outperforms existing methods, and under a Bayesian network model for extreme values with propagating noise, it is shown that the QTree estimator returns for n → ∞ a.s. the correct tree.
Tropical Support Vector Machines: Evaluations and Extension to Function Spaces
TLDR
It is shown theoretically by extreme value statistics that the tropical SVMs for classifying data points from two Gaussian distributions as well as empirical data sets of different neuron types are fairly robust against the curse of dimensionality.
Graphical models for extremes
TLDR
A general theory of conditional independence for multivariate Pareto distributions is introduced that enables the definition of graphical models and sparsity for extremes and it is shown that, similarly to the Gaussian case, the sparsity pattern of a general extremal graphical model can be read off from suitable inverse covariance matrices.

References

SHOWING 1-10 OF 49 REFERENCES
Graphical models for extremes
TLDR
A general theory of conditional independence for multivariate Pareto distributions is introduced that enables the definition of graphical models and sparsity for extremes and it is shown that, similarly to the Gaussian case, the sparsity pattern of a general extremal graphical model can be read off from suitable inverse covariance matrices.
Bayesian Networks for Max-Linear Models
TLDR
It is argued that the structure of a minimal network asymptotically can be identified completely from observational data.
Recursive max-linear models with propagating noise
TLDR
A probabilistic analysis of this new noisy model reveals that the unique minimum DAG representing the distribution of the non-noisy model remains unchanged and identifiable, and it is proved that the estimated Kleene star matrix converges to a matrix of independent Weibull entries after proper centering and scaling.
Identifying independence in bayesian networks
TLDR
This article characterizes all independence assertions that logically follow from the topology of a network and develops a linear time algorithm that identifies these assertions and is shown to work for a broad class of nonprobabilistic independencies.
Unifying Markov Properties for Graphical Models
Several types of graphs with different conditional independence interpretations --- also known as Markov properties --- have been proposed and used in graphical models. In this paper we unify these
Independence properties of directed markov fields
TLDR
A criterion for conditional independence of two groups of variables given a third is given and named as the directed, global Markov property and it is argued that this criterion is easy to use, it is sharper than that given by Kiiveri, Speed, and Carlin and equivalent to that of Pearl.
Max-linear models on directed acyclic graphs
TLDR
All max- linear models which are generated by a recursive structural equation model are characterized, and it is shown that its max-linear coefficient matrix is the solution of a fixed point equation.
Estimating an extreme Bayesian network via scalings
Equivalence and Synthesis of Causal Models
TLDR
The canonical representation presented here yields an efficient algorithm for determining when two embedded causal models reflect the same dependency information, which leads to a model theoretic definition of causation in terms of statistical dependencies.
...
...