Multilayer feedforward networks are universal approximators
The DOI matches its Crossref record; 16 of the 19 references it cites were found in a registry or an index.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Aligning with the theoretical results that MLPs with at least two layers are strictly more expressive than linear models (Hornik, Stinchcombe, and White 1989)”
From Titans: Learning to Memorize at Test Time (2025) · cites this paper as Hornik, Stinchcombe, and White 1989
In the citing paper
“Our chip-in-the-loop progressive fine-tuning overcomes nonlinear model errors by exploiting the intrinsic nonlinear universal approximation capacity of the deep neural network45, and furthermore eliminates the need for weight re-programming.”
From A compute-in-memory chip based on resistive random-access memory (2022) · cites this paper as [45]
In the citing paper
“deep neural networks (DNNs) which have the capacity to approximate any function, provided they are sufficiently complex and have sufficient examples to learn from.This capability is present regardless of whether they are composed of dense fully connected (Hornik et al. 1989) or convolutional layers (Zhou 2020).”
From Yield Prediction Through Integration of Genetic, Environment, and Management Data Through Deep Learning (2023) · cites this paper as Hornik et al. 1989
In the citing paper
“While feed-forward networks can in principle (given sufficient scale) learn any function (Hornik et al., 1989), including lookup tables of associations”
From Memory Layers at Scale (2024) · cites this paper as Hornik et al., 1989
In the citing paper
“These theorems guarantee that, provided enough free parameters, feedforward neural networks can approximate any continuous function on a closed and bounded subset of R n to an arbitrary degree of accuracy.”
From Continuous-variable quantum neural networks (2019) · cites this paper as [54]
4 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Approximation by superpositions of a sigmoidal function
Cybenko, G. (1989) · Mathematics of Control, Signals, and Systems · doi:10.1007/bf02551274 · Cited by 13,883
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., & WhiteH. (1989) · Neural Networks · doi:10.5555/70405.70408 · Cited by 9,081
PE, & Billingsley, P. (1996) · Journal of the American Statistical Association · doi:10.2307/2291440 · Cited by 6,481
Principles of mathematical analysis
Rudin, W. (1964) · UCLA - Biblioteca de Ciencias y Tecnologia Felix Morales Bueno · Cited by 5,300
Proceedings Of The International Joint Conference On Neural Networks 2003
(2003) · Proceedings of the International Joint Conference on Neural Networks, 2003. · doi:10.1109/ijcnn.2003.1223944 · Cited by 950
Bingham, N. H., & Billingsley, P. (1980) · Journal of the Royal Statistical Society. Series A (General) · doi:10.2307/2982142 · Cited by 455
Kolmogorov, A. N. (1963) · American Mathematical Society Translations: Series 2 · doi:10.1090/trans2/028/04 · Cited by 454
Theory of the Back Propagation Neural Network
Hecht-Nielsen, R. (1989) · International Joint Conference on Neural Networks (IJCNN) · Cited by 324
𝜀-entropy and 𝜀-capacity of sets in functional spaces
Kolmogorov, A. N. A. T. (1961) · Translations - American Mathematical Society/Translations · doi:10.1090/trans2/017/10 · Cited by 285
Abu‐Mostafa, Y. S. (1987) · Cited by 257
Research works printed but not linked: 1
- White, H., & Wooldridge, J. M. (in press). Some results for sieve estimation with dependent observations. In W. Barnett, J. Powell, & G. Tauchen (Eds.), Nonparametric and semi-parametric methods in econometrics and statistic. New York: Cambridge University Press.
Cited by
Modern Applied Statistics with S
Venables, W. N., & Ripley, B. D. (2002) · Statistics and Computing · doi:10.1007/978-0-387-21706-2 · Cited by 24,305
Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019) · Journal of Computational Physics · doi:10.1016/j.jcp.2018.10.045 · Cited by 20,384
“…work we take a different approach by employing deep neural networks and leverage their well known capability as universal function approximators [11] In this setting, we can directly tackle nonlinear problems without the need for committing to any prior assumptions, linearization, or local…”
BackgroundDeep learning in neural networks: an overview
Schmidhuber, J. (2015) · Neural Networks · doi:10.1016/j.neunet.2014.09.003 · Cited by 18,570
“Many practitioners found solace in a theorem (Kolmogorov, 1965a; Hecht-Nielsen, 1989; Hornik et al., 1989) stating that an NN with a single layer of enough hidden units can approximate any multivariate continous function with arbitrary accuracy.”
BackgroundApproximation by superpositions of a sigmoidal function
Cybenko, G. (1989) · Mathematics of Control, Signals, and Systems · doi:10.1007/bf02551274 · Cited by 13,883
A Comprehensive Survey on Graph Neural Networks
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Yu, P. S. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.2978386 · Cited by 11,000
“Universal approximation It is well known that multi-perceptron feedforward neural networks with one hidden layer can approximate any Borel measurable functions [105].”
Background
Metadata
- Authors
- Hornik, K., Stinchcombe, M., & White, H.
- Affiliations
- TU Wien (AT)
- University of California San Diego (US)
- Corresponding author
- White, Halbert
- ORCID
- Hornik, Kurt
- Venue
- Neural Networks
- Publisher
- Elsevier BV
- Published
- 1989-01
- Volume
- 2
- Issue
- 5
- Pages
- 359-366
- ISSN
- 0893-6080 (ISSN-L)
- OpenAlex
- W2137983211
- Semantic Scholar
- f22f6972e66bdd2e769fa64b0df0a13063c0c101
- Cited by
- 21,728
- References
- 19
- Citations by year
- 2026: 1377 · 2025: 1700 · 2024: 1582 · 2023: 1583 · 2022: 1422 · 2021: 1343 · 2020: 1188 · 2019: 1026 · 2018: 806 · 2017: 637 · 2016: 565 · 2015: 512 · 2014: 503 · 2013: 464 · 2012: 443
- FWCI
- 101.38
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- www.elsevier.com/tdm/userlicense/1.0/ (tdm)
- Locations
- Neural Networks (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and ELM
- Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
- Keywords
- multilayer perceptron, universal approximation, Borel measurable functions, hidden units, function approximation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Multilayer feedforward networks are universal approximators. https://app.cochat.ai/papers/10.1016/0893-6080(89)90020-8
- The paper
- Hornik, K., Stinchcombe, M., & White, H. (1989). Multilayer feedforward networks are universal approximators. Neural Networks. https://doi.org/10.1016/0893-6080(89)90020-8
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.