Joint Causal Inference from Observational and Experimental Datasets

We introduce Joint Causal Inference (JCI), a powerful formulation of causal discovery from multiple datasets that allows to jointly learn both the causal structure and targets of interventions from statistical independences in pooled data. Compared with existing constraint-based approaches for causal discovery from multiple data sets, JCI offers several advantages: it allows for several different types of interventions in a unified fashion, it can learn intervention targets, it systematically pools data across different datasets which improves the statistical power of independence tests, and most importantly, it improves on the accuracy and identifiability of the predicted causal relations. A technical complication that arises in JCI is the occurrence of faithfulness violations due to deterministic relations. We propose a simple but effective strategy for dealing with this type of faithfulness violations. We implement it in ACID, a determinism-tolerant extension of Ancestral Causal Inference (ACI) (Magliacane et al., 2016), a recently proposed logic-based causal discovery method that improves reliability of the output by exploiting redundant information in the data. We illustrate the benefits of JCI with ACID with an evaluation on a simulated dataset.

• Publication year

  2016

• Venue

  Unknown venue

• Publication date

  2016-11-30

• Fields of study

  Mathematics, Computer Science

• Identifiers
  arXiv 1611.10351
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Ancestral Causal Inference

  2016influential reference
• Towards Robust and Versatile Causal Discovery for Business Applications

  2016influential reference
• Causal inference by using invariant prediction: identification and confidence intervals

  2015influential reference
• Extended Conditional Independence and Applications in Causal Inference

  2015cited by this paper
• Constraint-based Causal Discovery: Conflict Resolution with Answer Set Programming

  2014influential reference
• Constraint-based causal discovery from multiple interventions over overlapping variable sets

  2014influential reference
• Large-scale genetic perturbations reveal regulatory networks and an abundance of gene-specific repressors.

  2014cited by this paper
• Data-driven covariate selection for nonparametric estimation of causal effects

  2013cited by this paper
• Cyclic Causal Discovery from Continuous Equilibrium Data

  2013cited by this paper
• Conservative independence-based causal structure learning in absence of adjacency faithfulness

  2012cited by this paper
• Learning from Mixture of Experimental Data: A Constraint-Based Approach

  2012cited by this paper
• Learning high-dimensional directed acyclic graphs with latent and selection variables

  2011cited by this paper
• Characterization and greedy learning of interventional Markov equivalence classes of directed acyclic graphs

  2011cited by this paper
• Causal discovery in multiple models from different experiments

  2010influential reference
• On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias

  2008influential reference
• A sufficient condition for pooling data

  2008influential reference
• Exact Bayesian structure learning from uncertain interventions

  2007influential reference
• Probabilistic Soft Interventions in Conditional Gaussian Networks

  2005cited by this paper
• Causation, Prediction, and Search

  2003influential reference
• Influence Diagrams for Causal Modelling and Inference

  2002cited by this paper
• Causal Discovery from Changes

  2001influential reference
• Causal Discovery from a Mixture of Experimental and Observational Data

  1999cited by this paper
• A Simple Constraint-Based Algorithm for Efficiently Mining Observational Databases for Causal Relationships

  1997cited by this paper
• Identifying independence in bayesian networks

  1990influential reference
• Conditional Independence in Statistical Theory

  1979cited by this paper
• Causality : Models , Reasoning , and Inference

  year unknowninfluential reference

• No Foundations without Foundations - Why semi-mechanistic models are essential for regulatory biology

  2025cites this paper
• Improved baselines for causal structure learning on interventional data

  2023cites this paper
• Design and Development of OCPDT by computing Path Probability by multiplying Probability differences of all Branch wise internal nodes

  2022cites this paper
• Design and Development of Optimal Causal Probability Decision Tree by ComputingPath probability of Internal Causality nodes

  2022cites this paper
• Perturbation graphs, invariant prediction and causal relations in psychology

  2021influential citation
• On the Completeness of Causal Discovery in the Presence of Latent Confounding with Tiered Background Knowledge

  2020cites this paper
• Sample Efficient Active Learning of Causal Trees

  2019cites this paper
• Improved Causal Discovery from Longitudinal Data Using a Mixture of DAGs

  2019cites this paper
• Characterization and Learning of Causal Graphs with Latent Variables from Soft Interventions

  2019cites this paper
• Switching Regression Models and Causal Inference in the Presence of Discrete Latent Variables

  2018cites this paper
• FASK with Interventional Knowledge Recovers Edges from the Sachs Model

  2018cites this paper
• Characterizing and Learning Equivalence Classes of Causal DAGs under Interventions

  2018cites this paper
• Personalized Network Modeling in Psychopathology: The Importance of Contemporaneous and Temporal Connections

  2018cites this paper
• Experimental Design for Learning Causal Graphs with Latent Variables

  2017cites this paper
• Logics for causal inference under uncertainty

  2017cites this paper
• Missing Data: A Preliminary Study

  year unknowncites this paper