Learn Before
  • Cast Cause-Effect Pair Problem as Machine Learning Classification Problem

Mother Distribution

The mother distribution is a distribution over join distributions of pairs of variables. The authors adopt the following definition of a mother distribution:

A mother distribution PM(PΠ=π(X,Y),G=g)P_M( P_{Π = π} ( X , Y ), G = g ) is a distribution over pairs (PΠ=π(X,Y),G=g)( P_{Π = π} ( X , Y ), G = g ), where ΠΠ and GG are random variables:

  • Π=πΠ = π indexes distribution PΠ=π(X,Y)P_{Π = π} ( X , Y ) over random variables XX and YY ; ππ may possibly take an infinite number of values (e.g. parameters of a generative process)
  • G=gG = g labels causal relationship, g{XY,XY,XY,XY}g ∈ \{ X → Y , X ← Y , X \leftrightarrow Y , X ⊥ Y \}.

0

1

5 years ago

Tags

Data Science

Related
  • Mother Distribution

  • Recast as a Two-Class Classification Problem

  • Definition of Pairwise Causal and Non-causal Relationship (in the book)

  • General Structure of Discriminant Learning Algorithms for Pairwise Causal Discovery

  • Limitations of the Framework

Learn After
  • Examples of Mother Distribution

  • Double Random Process for Dataset Generation