Journal article · 2024 · Open access

Marginal effects for non-linear prediction functions

Christian A. Scholbeck, Giuseppe Casalicchio, Christoph Molnar, Bernd Bischl, Christian Heumann

Data Mining and Knowledge Discovery, 38(5), 2997–3042 · Published 27 February 2024 · DOI 10.1007/s10618-023-00993-x

Free PDFView at the publisherCite this paper

In brief

“Marginal effects for non-linear prediction functions” is a 2024 journal article by Christian A. Scholbeck and 4 co-authors, published in Data Mining and Knowledge Discovery. OpenAlex counts 17 citations as of 16 November 2026. It is open access under a CC BY licence: the full text is free to read at Data Mining and Knowledge Discovery.

Citations
17
Most cited in
2024 (5)
Field-weighted impact
1.56
References
57
Authors
5

Abstract

Beta coefficients for linear regression models represent the ideal form of an interpretable feature effect. However, for non-linear models such as generalized linear models, the estimated coefficients cannot be interpreted as a direct feature effect on the predicted outcome. Hence, marginal effects are typically used as approximations for feature effects, either as derivatives of the prediction function or forward differences in prediction due to changes in feature values. While marginal effects are commonly used in many scientific fields, they have not yet been adopted as a general model-agnostic interpretation method for machine learning models. This may stem from the ambiguity surrounding marginal effects and their inability to deal with the non-linearities found in black box models. We introduce a unified definition of forward marginal effects (FMEs) that includes univariate and multivariate, as well as continuous, categorical, and mixed-type features. To account for the non-linearity of prediction functions, we introduce a non-linearity measure for FMEs. Furthermore, we argue against summarizing feature effects of a non-linear prediction function in a single metric such as the average marginal effect. Instead, we propose to average homogeneous FMEs within population subgroups, which serve as conditional feature effect estimates.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Scholbeck, C. A., Casalicchio, G., Molnar, C., Bischl, B., & Heumann, C. (2024). Marginal effects for non-linear prediction functions. Data Mining and Knowledge Discovery, 38(5), 2997–3042. https://doi.org/10.1007/s10618-023-00993-x

MLA 9

Scholbeck, Christian A., et al. “Marginal effects for non-linear prediction functions.” Data Mining and Knowledge Discovery, vol. 38, no. 5, 2024, pp. 2997–3042. https://doi.org/10.1007/s10618-023-00993-x.

Chicago

Scholbeck, Christian A., Giuseppe Casalicchio, Christoph Molnar, Bernd Bischl, and Christian Heumann. 2024. “Marginal effects for non-linear prediction functions.” Data Mining and Knowledge Discovery 38 (5): 2997–3042. https://doi.org/10.1007/s10618-023-00993-x.

Harvard

Scholbeck, C. A. et al. (2024) ‘Marginal effects for non-linear prediction functions’, Data Mining and Knowledge Discovery, 38(5), pp. 2997–3042. doi:10.1007/s10618-023-00993-x.

IEEE

C. A. Scholbeck, G. Casalicchio, C. Molnar, B. Bischl, and C. Heumann, “Marginal effects for non-linear prediction functions,” Data Mining and Knowledge Discovery, vol. 38, no. 5, pp. 2997–3042, 2024, doi: 10.1007/s10618-023-00993-x.

BibTeX

@article{scholbeck2024marginal,
  title     = {{Marginal effects for non-linear prediction functions}},
  author    = {Scholbeck, Christian A. and Casalicchio, Giuseppe and Molnar, Christoph and Bischl, Bernd and Heumann, Christian},
  journal   = {Data Mining and Knowledge Discovery},
  year      = {2024},
  volume    = {38},
  number    = {5},
  pages     = {2997--3042},
  doi       = {10.1007/s10618-023-00993-x},
  url       = {https://doi.org/10.1007/s10618-023-00993-x}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
202211
202311
202455
202555
202655

Its field-weighted citation impact is 1.56: it has been cited 1.56 times as often as the average work of the same field, type and year (1.00 is the world average).

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)

  • marginal effects
  • model interpretability
  • average marginal effects
  • subpopulations
  • black-box models
  • generalized linear models

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Scholbeck, C. A., Casalicchio, G., Molnar, C., Bischl, B., & Heumann, C. (2024). Marginal effects for non-linear prediction functions. Data Mining and Knowledge Discovery, 38(5), 2997–3042. https://doi.org/10.1007/s10618-023-00993-x

How many times has this paper been cited?

17 times, according to OpenAlex on 16 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access: read the free full text at Data Mining and Knowledge Discovery.

Who wrote this paper?

Christian A. Scholbeck, Giuseppe Casalicchio, Christoph Molnar, Bernd Bischl, Christian Heumann.

What is the DOI of this paper?

10.1007/s10618-023-00993-x. The permanent link is https://doi.org/10.1007/s10618-023-00993-x.

Bibliographic data from OpenAlex (CC0), last checked 16 November 2026. ResearchGuru does not host this paper; the links go to the publisher or the repository that does.