Contents
Why are Bayesian networks used for causal inference?
Bayesian Networks offer a graphical representation that is reasonably interpretable and easily explainable; Relationships captured between variables in a Bayesian Network are more complex yet hopefully more informative than a conventional model;
What do you need to know about Bayesian networks?
Bayesian Networks are probabilistic graphical models that represent the dependency structure of a set of variables and their joint distribution efficiently in a factorised way.
Which is the best framework for causal inference?
Potential outcomes framework (Rubin causal model) , propensity score matching and structural causal models are, arguably, the most popular frameworks for observational causal inference. Here, we focus on the structural causal models and one particular type, Bayesian Networks.
Which is the Central class for causal inference?
The BayesianNetwork class is the central class for the causal inference analysis in the package. It is built on top of the StructureModel, which is an extension of networkx.DiGraph StructureModel represents a causal graph, a DAG of the respective BN and holds directed edges, describing a cause -> effect relationship.
How are probabilistic graphical models used in causal inference?
Probabilistic graphical models: principles and techniques by D. Koller and N. Friedman. Structural causal models represent causal dependencies using graphical models that provide an intuitive visualisation by representing variables as nodes and relationships between variables as edges in a graph.
Why are correlations so important in causal inference?
One possible explanation for correlation between variables where neither causes the other is the presence of confounding variables that influence both the target and a driver of that target. Unobserved confounding variables are severe threats when doing causal inference on observational data.