Measurement dependence inducing latent causal models

Measurement dependence inducing latent causal models

Abstract

We consider the task of causal structure learning over measurement dependence inducing latent (MeDIL) causal models. We show that thistask can be framed in terms of the graph theoretic problem of finding edge clique covers, resulting in an algorithm for returning minimalMeDIL causal models (minMCMs). This algorithm is non-parametric, requiring no assumptions about linearity or Gaussianity. Furthermore, despite rather weak assumptions aboutthe class of MeDIL causal models, we show that minimality in minMCMs implies some rather specific and interesting properties. By establishing MeDIL causal models as a semantics for edge clique covers, we also provide a starting point for future work further connecting causal structure learning to developments in graph theory and network science.

Grafik Top
Authors
  • Markham, Alex
  • Grosse-Wentrup, Moritz
Grafik Top
Shortfacts
Category
Paper in Conference Proceedings or in Workshop Proceedings (Paper)
Event Title
36th Conference in Uncertainty in Artificial Intelligence (UAI) 2020
Divisions
Neuroinformatics
Subjects
Kuenstliche Intelligenz
Event Location
virtual event
Event Type
Conference
Event Dates
3-6 Aug 2020
Series Name
Proceedings of Machine Learning Research (PMLR)
Publisher
Proceedings of the 36th Conference on Uncertainty in Artificial Intelligence (UAI), Proceedings of Machine Learning Research
Page Range
pp. 590-599
Date
August 2020
Export
Grafik Top