Targeted Causal Elicitation

Nazaal Ibrahim; Ti John; Zhigao Guo; Samuel Kaski

Targeted Causal Elicitation

Nazaal Ibrahim^*, Ti John, Zhigao Guo, Samuel Kaski

^*Corresponding author for this work

University of Manchester

Research output: Contribution to conference › Paper › Scientific › peer-review

Abstract

We look at the problem of learning causal structure for a fixed downstream causal effect optimization task. In contrast to previous work which often focuses on running interventional experiments, we consider an often overlooked source of information - the domain expert. In the Bayesian setting, this amounts to augmenting the likelihood with a user model whose parameters account for possible biases of the expert. Such a model can allow for active elicitation in a manner that is most informative to the optimization task at hand.

Original language	English
Publication status	Published - 2022
MoE publication type	Not Eligible
Event	NeurIPS Workshop on Causality for Real-world Impact - New Orleans, United States Duration: 2 Dec 2022 → 2 Dec 2022 https://www.cml-4-impact.vanderschaar-lab.com/

Workshop

Workshop	NeurIPS Workshop on Causality for Real-world Impact
Abbreviated title	CML4Impact
Country/Territory	United States
City	New Orleans
Period	02/12/2022 → 02/12/2022
Internet address	https://www.cml-4-impact.vanderschaar-lab.com/

Access to Document

https://openreview.net/pdf?id=oqRw-a4rf36

Cite this

@conference{07500236550647e1b2477f0d67df9d9d,

title = "Targeted Causal Elicitation",

abstract = "We look at the problem of learning causal structure for a fixed downstream causal effect optimization task. In contrast to previous work which often focuses on running interventional experiments, we consider an often overlooked source of information - the domain expert. In the Bayesian setting, this amounts to augmenting the likelihood with a user model whose parameters account for possible biases of the expert. Such a model can allow for active elicitation in a manner that is most informative to the optimization task at hand.",

author = "Nazaal Ibrahim and Ti John and Zhigao Guo and Samuel Kaski",

year = "2022",

language = "English",

note = "NeurIPS Workshop on Causality for Real-world Impact, CML4Impact ; Conference date: 02-12-2022 Through 02-12-2022",

url = "https://www.cml-4-impact.vanderschaar-lab.com/",

}

TY - CONF

T1 - Targeted Causal Elicitation

AU - Ibrahim, Nazaal

AU - John, Ti

AU - Guo, Zhigao

AU - Kaski, Samuel

PY - 2022

Y1 - 2022

N2 - We look at the problem of learning causal structure for a fixed downstream causal effect optimization task. In contrast to previous work which often focuses on running interventional experiments, we consider an often overlooked source of information - the domain expert. In the Bayesian setting, this amounts to augmenting the likelihood with a user model whose parameters account for possible biases of the expert. Such a model can allow for active elicitation in a manner that is most informative to the optimization task at hand.

AB - We look at the problem of learning causal structure for a fixed downstream causal effect optimization task. In contrast to previous work which often focuses on running interventional experiments, we consider an often overlooked source of information - the domain expert. In the Bayesian setting, this amounts to augmenting the likelihood with a user model whose parameters account for possible biases of the expert. Such a model can allow for active elicitation in a manner that is most informative to the optimization task at hand.

M3 - Paper

T2 - NeurIPS Workshop on Causality for Real-world Impact

Y2 - 2 December 2022 through 2 December 2022

ER -

Targeted Causal Elicitation

Abstract

Workshop

Access to Document

Fingerprint

Cite this