Cookies?
Library Header Image
LSE Research Online LSE Library Services

Counterfactuals for the future

Bynum, Lucius E.J., Loftus, Joshua R. ORCID: 0000-0002-2905-1632 and Stoyanovich, Julia (2023) Counterfactuals for the future. In: Williams, Brian, Chen, Yiling and Neville, Jennifer, (eds.) AAAI-23 Special Tracks. Proceedings of the 37th AAAI Conference on Artificial Intelligence, AAAI 2023. AAAI Press, pp. 14144-14152. ISBN 9781577358800

Full text not available from this repository.

Abstract

Counterfactuals are often described as ‘retrospective,’ focusing on hypothetical alternatives to a realized past. This description relates to an often implicit assumption about the structure and stability of exogenous variables in the system being modeled — an assumption that is reasonable in many settings where counterfactuals are used. In this work, we consider cases where we might reasonably make a different assumption about exogenous variables; namely, that the exogenous noise terms of each unit do exhibit some unit-specific structure and/or stability. This leads us to a different use of counterfactuals — a forward-looking rather than retrospective counterfactual. We introduce “counterfactual treatment choice,” a type of treatment choice problem that motivates using forward-looking counterfactuals. We then explore how mismatches between interventional versus forward-looking counterfactual approaches to treatment choice, consistent with different assumptions about exogenous noise, can lead to counterintuitive results.

Item Type: Book Section
Official URL: https://ojs.aaai.org/index.php/AAAI/index
Additional Information: © 2023, Association for the Advancement of Artificial Intelligence (www.aaai.org).
Divisions: Statistics
Date Deposited: 01 Sep 2023 10:24
Last Modified: 26 May 2024 05:09
URI: http://eprints.lse.ac.uk/id/eprint/120115

Actions (login required)

View Item View Item