Associate Professor in Machine Learning

Compositional Abstraction Error

less than 1 minute read

Published:

In previous posts we have seen how to define abstractions between causal models, and how to compute automatically the abstraction error when switching from a micromodel to a macromodel.

In this notebook we consider an important theoretic property of abstraction, that is, composition of abstractions. Specifically, we re-implement the examples provided in a paper on the compositionally of abstraction, and we empirically show that a proper choice of the metric to estimate the abstraction error guarantees a bound on the abstraction error of a composed abstraction.

Notebook here

You May Also Enjoy

Reviewing Consistencies

less than 1 minute read

Published:

In previous posts we have discussed theoretical and conceptual properties of abstractions between causal models, while at the same time implementing code for checking our statements and for running simulations.

Scenarios of Confounding

less than 1 minute read

Published:

Causality and causal inference deal with expressing and reasoning about relationships of cause and effects, and structural causal model provide a rigorous formalism to assess causality.