Causal Inference Using the Algorithmic Markov Condition

  title={Causal Inference Using the Algorithmic Markov Condition},
  author={Dominik Janzing and Bernhard Sch{\"o}lkopf},
  journal={IEEE Transactions on Information Theory},
Inferring the causal structure that links n observables is usually based upon detecting statistical dependences and choosing simple graphs that make the joint measure Markovian. Here we argue why causal inference is also possible when the sample size is one. We develop a theory how to generate causal graphs explaining similarities between single objects. To this end, we replace the notion of conditional stochastic independence in the causal Markov condition with the vanishing of conditional… 

Restricted structural equation models for causal inference

This paper focuses on causal inference, which tries to solve the following problem: given i.d. data from a joint distribution, one tries to infer the underlying causal DAG (directed acyclic graph), in which each node represents one of the observed variables.

Replacing Causal Faithfulness with Algorithmic Independence of Conditionals

This paper compares IC with causal faithfulness (FF), stating that only those conditional independences that are implied by the causal Markov condition hold true.

Causal Markov Condition for Submodular Information Measures

This work forms a generalized causal Markov condition (CMC) for any kind of observations on which independence is defined via an arbitrary submodular information measure and shows that this CMC is justified if one restricts the attention to a class of causal mechanisms that are adapted to the respective information measure.

Novel Methods of Causal Inference and their impact for standard prediction tasks

  • Computer Science
  • 2012
A causal inference scenario where the nodes of the causal DAG need not be random variables, but arbitrary mathematical objects x1, . . . , xn that formalize observations, and it is argued that dependences between xi and xj can be defined by any information measure R as the difference R(xi) + R(xj) − R (xi, xj), provided that the definition of R guarantees non-negativity of this expression.

Causal Discovery Beyond Conditional Independences

This thesis addresses the problem of causal discovery, that is, recovering the underlying causal structure based on the joint probability distribution of the observed random variables, and estimates a set of Markov equivalent graphs.

Distinguishing Cause and Effect via Second Order Exponential Models

This work proposes a method to infer causal structures containing both discrete and continuous variables and describes that sometimes a causal hypothesis must be rejected because P(effect|cause) and P(cause) share algorithmic information (which is untypical if they are chosen independently).

Identifiability of Causal Graphs using Functional Models

A main theorem is proved that if the data generating process belongs to an IFMOC, one can identify the complete causal graph and is the first identifiability result of this kind that is not limited to linear functional relationships.

Telling cause from effect by local and global regression

The linear-time Slope and Sloper algorithms are introduced that through thorough empirical evaluation the authors show outperform the state of the art by a wide margin.

Discovering Fully Oriented Causal Networks

The GLOBE algorithm is introduced, which greedily adds, removes, and orients edges such that it minimizes the overall cost, and performs very well in practice, beating the state of the art by a margin.

Inferring latent structures via information inequalities

An information-theoretic approach is proposed, based on the insight that conditions on entropies of Bayesian networks take the form of simple linear inequalities, and an algorithm for deriving entropic tests for latent structures is described.



Inference of Graphical Causal Models: Representing the Meaningful Information of Probability Distributions

It is argued that if the shortest description of the joint distribution is given by separate descriptions of the conditional distributions for each variable given its effects, the decomposition given by the DAG should be considered as the top-ranked causal hypothesis.

Causal Inference by Choosing Graphs with Most Plausible Markov Kernels

This paper discusses the inference rule for causal relationships between two variables in detail, applies it to a real-world temperature data set with known causality and shows that the method provides a correct result for the example.

Causation, prediction, and search

The authors axiomatize the connection between causal structure and probabilistic independence, explore several varieties of causal indistinguishability, formulate a theory of manipulation, and develop asymptotically reliable procedures for searching over equivalence classes of causal models.

Causal reasoning by evaluating the complexity of conditional densities with kernel methods

A Bayesian Approach to Causal Discovery

The general Bayesian approach to causal discovery is described and approximation methods for missing data and hidden variables are reviewed, and differences between the Bayesian and constraint-based methods are illustrated using artificial and real examples.

On causally asymmetric versions of Occam's Razor and their relation to thermodynamics

In real-life statistical data, it seems that conditional probabilities for the effect given their causes tend to be less complex and smoother than conditionals for causes, given their effects. We

Causal Models as Minimal Descriptions of Multivariate Systems

By applying the minimality principle for model selection, one should seek the model that describes the data by a code of minimal length, in order to build a model containing the meaningful information.

Algorithmic statistics

The algorithmic theory of statistic, sufficient statistic, and minimal sufficient statistic is developed and it is shown that a function is a probabilistic sufficient statistic iff it is with high probability (in an appropriate sense) an algorithmic sufficient statistic.

Causal Inference Using Nonnormality

Path analysis, often applied to observational data to study causal structures, describes causal relationship between observed variables. The path analysis is of confirmatory nature and can make

On Universal Prediction and Bayesian Confirmation