Distinguishing cause from effect based on exogeneity

Kun ZHANG, Jiji ZHANG, Bernhard SCHÖKOPF

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

Abstract

Recent developments in structural equation modeling have produced several methods that can usually distinguish cause from effect in the two-variable case. For that purpose, how- ever, one has to impose substantial structural constraints or smoothness assumptions on the functional causal mod- els. In this paper, we consider the problem of determining the causal direction from a related but different point of view, and propose a new framework for causal direction de- termination. We show that it is possible to perform causal inference based on the condition that the cause is “exoge- nous” for the parameters involved in the generating pro- cess from the cause to the effect. In this way, we avoid the structural constraints required by the SEM-based ap- proaches. In particular, we exploit nonparametric methods to estimate marginal and conditional distributions, and pro- pose a bootstrap-based approach to test for the exogeneity condition; the testing results indicate the causal direction between two variables. The proposed method is validated on both synthetic and real data.
Original languageEnglish
Title of host publicationProceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge
EditorsR. RAMANUJAM
PublisherTARK 2015
Pages261-271
Number of pages11
ISBN (Print)9780692461389
Publication statusPublished - 22 Apr 2015
EventFifteenth conference on Theoretical Aspects of Rationality and Knowledge - Carnegie Mellon University, Pittsburgh, United States
Duration: 4 Jun 20156 Jun 2015
https://www.imsc.res.in/tark/tark15.html

Conference

ConferenceFifteenth conference on Theoretical Aspects of Rationality and Knowledge
Abbreviated titleTARK 2015
CountryUnited States
CityPittsburgh
Period4/06/156/06/15
OtherTARK 2015 took place at the Carnegie Mellon University, Pittsburgh, USA during June 4--6, 2015.
The mission of the TARK conferences is to bring together researchers from a wide variety of fields, including Artificial Intelligence, Cryptography, Distributed Computing, Economics and Game Theory, Linguistics, Philosophy, and Psychology, in order to further our understanding of interdisciplinary issues involving reasoning about rationality and knowledge.
Internet address

Fingerprint

Scanning electron microscopy
Testing

Keywords

  • Causal discovery
  • causal direction
  • exogeneity
  • statistical independence
  • bootstrap

Cite this

ZHANG, K., ZHANG, J., & SCHÖKOPF, B. (2015). Distinguishing cause from effect based on exogeneity. In R. RAMANUJAM (Ed.), Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge (pp. 261-271). TARK 2015.
ZHANG, Kun ; ZHANG, Jiji ; SCHÖKOPF, Bernhard. / Distinguishing cause from effect based on exogeneity. Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge. editor / R. RAMANUJAM. TARK 2015, 2015. pp. 261-271
@inproceedings{4459cf1c6e814d7d88900f6196077a3c,
title = "Distinguishing cause from effect based on exogeneity",
abstract = "Recent developments in structural equation modeling have produced several methods that can usually distinguish cause from effect in the two-variable case. For that purpose, how- ever, one has to impose substantial structural constraints or smoothness assumptions on the functional causal mod- els. In this paper, we consider the problem of determining the causal direction from a related but different point of view, and propose a new framework for causal direction de- termination. We show that it is possible to perform causal inference based on the condition that the cause is “exoge- nous” for the parameters involved in the generating pro- cess from the cause to the effect. In this way, we avoid the structural constraints required by the SEM-based ap- proaches. In particular, we exploit nonparametric methods to estimate marginal and conditional distributions, and pro- pose a bootstrap-based approach to test for the exogeneity condition; the testing results indicate the causal direction between two variables. The proposed method is validated on both synthetic and real data.",
keywords = "Causal discovery, causal direction, exogeneity, statistical independence, bootstrap",
author = "Kun ZHANG and Jiji ZHANG and Bernhard SCH{\"O}KOPF",
year = "2015",
month = "4",
day = "22",
language = "English",
isbn = "9780692461389",
pages = "261--271",
editor = "R. RAMANUJAM",
booktitle = "Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge",
publisher = "TARK 2015",

}

ZHANG, K, ZHANG, J & SCHÖKOPF, B 2015, Distinguishing cause from effect based on exogeneity. in R RAMANUJAM (ed.), Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge. TARK 2015, pp. 261-271, Fifteenth conference on Theoretical Aspects of Rationality and Knowledge, Pittsburgh, United States, 4/06/15.

Distinguishing cause from effect based on exogeneity. / ZHANG, Kun; ZHANG, Jiji; SCHÖKOPF, Bernhard.

Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge. ed. / R. RAMANUJAM. TARK 2015, 2015. p. 261-271.

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

TY - GEN

T1 - Distinguishing cause from effect based on exogeneity

AU - ZHANG, Kun

AU - ZHANG, Jiji

AU - SCHÖKOPF, Bernhard

PY - 2015/4/22

Y1 - 2015/4/22

N2 - Recent developments in structural equation modeling have produced several methods that can usually distinguish cause from effect in the two-variable case. For that purpose, how- ever, one has to impose substantial structural constraints or smoothness assumptions on the functional causal mod- els. In this paper, we consider the problem of determining the causal direction from a related but different point of view, and propose a new framework for causal direction de- termination. We show that it is possible to perform causal inference based on the condition that the cause is “exoge- nous” for the parameters involved in the generating pro- cess from the cause to the effect. In this way, we avoid the structural constraints required by the SEM-based ap- proaches. In particular, we exploit nonparametric methods to estimate marginal and conditional distributions, and pro- pose a bootstrap-based approach to test for the exogeneity condition; the testing results indicate the causal direction between two variables. The proposed method is validated on both synthetic and real data.

AB - Recent developments in structural equation modeling have produced several methods that can usually distinguish cause from effect in the two-variable case. For that purpose, how- ever, one has to impose substantial structural constraints or smoothness assumptions on the functional causal mod- els. In this paper, we consider the problem of determining the causal direction from a related but different point of view, and propose a new framework for causal direction de- termination. We show that it is possible to perform causal inference based on the condition that the cause is “exoge- nous” for the parameters involved in the generating pro- cess from the cause to the effect. In this way, we avoid the structural constraints required by the SEM-based ap- proaches. In particular, we exploit nonparametric methods to estimate marginal and conditional distributions, and pro- pose a bootstrap-based approach to test for the exogeneity condition; the testing results indicate the causal direction between two variables. The proposed method is validated on both synthetic and real data.

KW - Causal discovery

KW - causal direction

KW - exogeneity

KW - statistical independence

KW - bootstrap

UR - https://arxiv.org/abs/1504.05651

M3 - Conference paper (refereed)

SN - 9780692461389

SP - 261

EP - 271

BT - Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge

A2 - RAMANUJAM, R.

PB - TARK 2015

ER -

ZHANG K, ZHANG J, SCHÖKOPF B. Distinguishing cause from effect based on exogeneity. In RAMANUJAM R, editor, Proceedings of the 15th conference on Theoretical Aspects of Rationality and Knowledge. TARK 2015. 2015. p. 261-271