We formulate a general framework for building structural causal models (SCMs)
with deep learning components. The proposed approach employs normalising flows
and variational inference to enable tractable inference of exogenous noise
variables - a crucial step for counterfactual inference that is missing from
existing deep causal learning methods. Our framework is validated on a
synthetic dataset built on MNIST as well as on a real-world medical dataset of
brain MRI scans. Our experimental results indicate that we can successfully
train deep SCMs that are capable of all three levels of Pearl's ladder of
causation: association, intervention, and counterfactuals, giving rise to a
powerful new approach for answering causal questions in imaging applications
and beyond. The code for all our experiments is available at
https://github.com/biomedia-mira/deepscm.
Description
[2006.06485] Deep Structural Causal Models for Tractable Counterfactual Inference
%0 Journal Article
%1 pawlowski2020structural
%A Pawlowski, Nick
%A Castro, Daniel C.
%A Glocker, Ben
%D 2020
%K causal-analysis neurips2020 readings
%T Deep Structural Causal Models for Tractable Counterfactual Inference
%U http://arxiv.org/abs/2006.06485
%X We formulate a general framework for building structural causal models (SCMs)
with deep learning components. The proposed approach employs normalising flows
and variational inference to enable tractable inference of exogenous noise
variables - a crucial step for counterfactual inference that is missing from
existing deep causal learning methods. Our framework is validated on a
synthetic dataset built on MNIST as well as on a real-world medical dataset of
brain MRI scans. Our experimental results indicate that we can successfully
train deep SCMs that are capable of all three levels of Pearl's ladder of
causation: association, intervention, and counterfactuals, giving rise to a
powerful new approach for answering causal questions in imaging applications
and beyond. The code for all our experiments is available at
https://github.com/biomedia-mira/deepscm.
@article{pawlowski2020structural,
abstract = {We formulate a general framework for building structural causal models (SCMs)
with deep learning components. The proposed approach employs normalising flows
and variational inference to enable tractable inference of exogenous noise
variables - a crucial step for counterfactual inference that is missing from
existing deep causal learning methods. Our framework is validated on a
synthetic dataset built on MNIST as well as on a real-world medical dataset of
brain MRI scans. Our experimental results indicate that we can successfully
train deep SCMs that are capable of all three levels of Pearl's ladder of
causation: association, intervention, and counterfactuals, giving rise to a
powerful new approach for answering causal questions in imaging applications
and beyond. The code for all our experiments is available at
https://github.com/biomedia-mira/deepscm.},
added-at = {2020-09-28T00:50:40.000+0200},
author = {Pawlowski, Nick and Castro, Daniel C. and Glocker, Ben},
biburl = {https://www.bibsonomy.org/bibtex/253ce714a66071a881de2880e9645e0a9/kirk86},
description = {[2006.06485] Deep Structural Causal Models for Tractable Counterfactual Inference},
interhash = {14f9039c923b636bb831eda1564ed39a},
intrahash = {53ce714a66071a881de2880e9645e0a9},
keywords = {causal-analysis neurips2020 readings},
note = {cite arxiv:2006.06485},
timestamp = {2020-09-28T00:50:40.000+0200},
title = {Deep Structural Causal Models for Tractable Counterfactual Inference},
url = {http://arxiv.org/abs/2006.06485},
year = 2020
}