IDEAS home Printed from https://ideas.repec.org/a/gam/jmathe/v12y2024i5p696-d1347243.html
   My bibliography  Save this article

Invariant Feature Learning Based on Causal Inference from Heterogeneous Environments

Author

Listed:
  • Hang Su

    (School of Mathematics, Renmin University of China, Beijing 100872, China)

  • Wei Wang

    (School of Mathematics, Renmin University of China, Beijing 100872, China)

Abstract

Causality has become a powerful tool for addressing the out-of-distribution (OOD) generalization problem, with the idea of invariant causal features across domains of interest. Most existing methods for learning invariant features are based on optimization, which typically fails to converge to the optimal solution. Therefore, obtaining the variables that cause the target outcome through a causal inference method is a more direct and effective method. This paper presents a new approach for invariant feature learning based on causal inference (IFCI). IFCI detects causal variables unaffected by the environment through the causal inference method. IFCI focuses on partial causal relationships to work efficiently even in the face of high-dimensional data. Our proposed causal inference method can accurately infer causal effects even when the treatment variable has more complex values. Our method can be viewed as a pretreatment of data to filter out variables whose distributions change between different environments, and it can then be combined with any learning method for classification and regression. The result of empirical studies shows that IFCI can detect and filter out environmental variables affected by the environment. After filtering out environmental variables, even a model with a simple structure and common loss function can have strong OOD generalization capability. Furthermore, we provide evidence to show that classifiers utilizing IFCI achieve higher accuracy in classification compared to existing OOD generalization algorithms.

Suggested Citation

  • Hang Su & Wei Wang, 2024. "Invariant Feature Learning Based on Causal Inference from Heterogeneous Environments," Mathematics, MDPI, vol. 12(5), pages 1-23, February.
  • Handle: RePEc:gam:jmathe:v:12:y:2024:i:5:p:696-:d:1347243
    as

    Download full text from publisher

    File URL: https://www.mdpi.com/2227-7390/12/5/696/pdf
    Download Restriction: no

    File URL: https://www.mdpi.com/2227-7390/12/5/696/
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Donald B. Rubin, 2005. "Causal Inference Using Potential Outcomes: Design, Modeling, Decisions," Journal of the American Statistical Association, American Statistical Association, vol. 100, pages 322-331, March.
    2. Jonas Peters & Peter Bühlmann & Nicolai Meinshausen, 2016. "Causal inference by using invariant prediction: identification and confidence intervals," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 78(5), pages 947-1012, November.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Hang Su & Wei Wang, 2023. "An Out-of-Distribution Generalization Framework Based on Variational Backdoor Adjustment," Mathematics, MDPI, vol. 12(1), pages 1-21, December.
    2. Dominik Rothenhäusler & Nicolai Meinshausen & Peter Bühlmann & Jonas Peters, 2021. "Anchor regression: Heterogeneous data meet causality," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 83(2), pages 215-246, April.
    3. Salvatore Bimonte & Antonella D’Agostino, 2021. "Tourism development and residents’ well-being: Comparing two seaside destinations in Italy," Tourism Economics, , vol. 27(7), pages 1508-1525, November.
    4. Lenz, Gabriel & Sahn, Alexander, 2017. "Achieving Statistical Significance with Covariates and without Transparency," MetaArXiv s42ba_v1, Center for Open Science.
    5. Sahar Saeed & Erica E. M. Moodie & Erin C. Strumpf & Marina B. Klein, 2018. "Segmented generalized mixed effect models to evaluate health outcomes," International Journal of Public Health, Springer;Swiss School of Public Health (SSPH+), vol. 63(4), pages 547-551, May.
    6. Fangting Zhou & Kejun He & Yang Ni, 2023. "Individualized causal discovery with latent trajectory embedded Bayesian networks," Biometrics, The International Biometric Society, vol. 79(4), pages 3191-3202, December.
    7. Yiran Zhang & Andrew Ying & Steve Edland & Lon White & Ronghui Xu, 2024. "Marginal Structural Illness-Death Models for Semi-competing Risks Data," Statistics in Biosciences, Springer;International Chinese Statistical Association, vol. 16(3), pages 668-692, December.
    8. Manuel S. González Canché, 2017. "Financial Benefits of Rapid Student Loan Repayment: An Analytic Framework Employing Two Decades of Data," The ANNALS of the American Academy of Political and Social Science, , vol. 671(1), pages 154-182, May.
    9. Almer, Christian & Winkler, Ralph, 2017. "Analyzing the effectiveness of international environmental policies: The case of the Kyoto Protocol," Journal of Environmental Economics and Management, Elsevier, vol. 82(C), pages 125-151.
    10. Sanford C. Gordon & Hannah K. Simpson, 2020. "Causes, theories, and the past in political science," Public Choice, Springer, vol. 185(3), pages 315-333, December.
    11. Angelov, Nikolay & Eliason, Marcus, 2014. "The effects of targeted labour market programs for job seekers with occupational disabilities," Working Paper Series 2014:27, IFAU - Institute for Evaluation of Labour Market and Education Policy.
    12. Mark Kattenberg & Bas Scheer & Jurre Thiel, 2023. "Causal forests with fixed effects for treatment effect heterogeneity in difference-in-differences," CPB Discussion Paper 452, CPB Netherlands Bureau for Economic Policy Analysis.
    13. Slutskin, L., 2017. "Graphical Statistical Methods for Studying Causal Effects. Bayesian Networks," Journal of the New Economic Association, New Economic Association, vol. 36(4), pages 12-30.
    14. Tianmeng Lyu & Björn Bornkamp & Guenther Mueller‐Velten & Heinz Schmidli, 2023. "Bayesian inference for a principal stratum estimand on recurrent events truncated by death," Biometrics, The International Biometric Society, vol. 79(4), pages 3792-3802, December.
    15. Riukula, Krista & Väänänen, Touko, 2024. "Estimating the Labour Market Impacts of Transport Projects in Finland," ETLA Working Papers 120, The Research Institute of the Finnish Economy.
    16. Nicholas Illenberger & Dylan S. Small & Pamela A. Shaw, 2019. "Regression to the Mean's Impact on the Synthetic Control Method: Bias and Sensitivity Analysis," Papers 1909.04706, arXiv.org.
    17. Silvana Tiedemann & Jorge Sanchez Canales & Felix Schur & Raffaele Sgarlato & Lion Hirth & Oliver Ruhnau & Jonas Peters, 2024. "Identifying Elasticities in Autocorrelated Time Series Using Causal Graphs," Papers 2409.15530, arXiv.org.
    18. Gary Henry & Roderick Rose & Doug Lauen, 2014. "Are value-added models good enough for teacher evaluations? Assessing commonly used models with simulated and actual data," Investigaciones de Economía de la Educación volume 9, in: Adela García Aracil & Isabel Neira Gómez (ed.), Investigaciones de Economía de la Educación 9, edition 1, volume 9, chapter 20, pages 383-405, Asociación de Economía de la Educación.
    19. Steven M. Smith, 2019. "The Relative Economic Merits of Alternative Water Rights," Working Papers 2019-08, Colorado School of Mines, Division of Economics and Business.
    20. Christian Almer & Ralph Winkler, 2012. "The Effect of Kyoto Emission Targets on Domestic CO2 Emissions: A Synthetic Control Approach," Diskussionsschriften dp1202, Universitaet Bern, Departement Volkswirtschaft.

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gam:jmathe:v:12:y:2024:i:5:p:696-:d:1347243. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: MDPI Indexing Manager (email available below). General contact details of provider: https://www.mdpi.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.