Current browse context:
stat.ML
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: Learning Representations for Counterfactual Inference
(Submitted on 12 May 2016 (this version), latest version 6 Jun 2018 (v3))
Abstract: Observational studies are rising in importance due to the widespread accumulation of data in fields such as healthcare, education, employment and ecology. We consider the task of answering counterfactual questions such as, "Would this patient have lower blood sugar had she received a different medication?". We propose a new algorithmic framework for counterfactual inference which brings together ideas from domain adaptation and representation learning. In addition to a theoretical justification, we perform an empirical comparison with previous approaches to causal inference from observational data. Our deep learning algorithm significantly outperforms the previous state-of-the-art.
Submission history
From: Fredrik D. Johansson [view email][v1] Thu, 12 May 2016 02:59:40 GMT (193kb,D)
[v2] Wed, 8 Jun 2016 17:04:07 GMT (194kb,D)
[v3] Wed, 6 Jun 2018 13:00:53 GMT (195kb,D)
Link back to: arXiv, form interface, contact.