On the finite sample complexity of causal discovery and the value of domain expertise

Causal discovery methods seek to identify causal relations between random variables from purely observational data, as opposed to actively collected experimental data where an experimenter intervenes on a subset of correlates. One of the seminal works in this area is the Inferred Causation (IC) algorithm, which guarantees successful causal discovery under the assumption of a conditional independence (CI) oracle: an oracle that can states whether two random variables are conditionally independent given another set of random variables. Practical implementations of this algorithm incorporate statistical tests for conditional independence, in place of a CI oracle. In this thesis, we analyze the sample complexity of causal discovery algorithms without a CI oracle: given a certain level of confidence, how many data points are needed for a causal discovery algorithm to identify a causal structure? Furthermore, our methods allow us to quantify the value of domain expertise in terms of data samples. Finally, we demonstrate the accuracy of these sample rates with numerical examples, and quantify the benefits of three types of domain expertise: sparsity priors, known causal directions, and known conditional dependencies.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Mechanical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2021

Author dc:creator

• Wadhwa, Samir

Contributors dc:contributor

• Dong, Roy
• Dullerud, Geir Eirik

• causality
• discrete Bayesian networks
• conditional independence testing
• family-wise error rate

Statement dc:rights

• Copyright 2021 Samir Wadhwa

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/110577

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/110577