三类命名结论对应原始论文 + 海外引用文献完整清单

📝 导读说明:本文献清单中的每篇文章均已生成独立的 Markdown 文件进行详细介绍。您可以点击每篇文献下方的“📄 查看本篇独立 Markdown 介绍”链接查看。

一、原始三篇奠基论文(陈天平三篇原文,所有命名的本源)

论文 1(构造性 Cybenko 定理,1990)

Chen T, Chen H, Liu R W. A Constructive Proof and An Extension of Cybenko's Approximation Theorem[C]//Computing Science and Statistics. Springer, 1992: 163-168.

📄 查看本篇独立 Markdown 介绍

论文 2(算子万能逼近、TW 激活判定,1995 IEEE TNN,核心命名来源)

Chen T, Chen H. Universal Approximation to Nonlinear Operators by Neural Networks with Arbitrary Activation Functions and Its Application to Dynamical Systems[J]. IEEE Transactions on Neural Networks, 1995, 6(4): 911-917.
DOI:10.1109/72.392252

📄 查看本篇独立 Markdown 介绍

论文 3(Hilbert 空间泛函逼近,1994 科学通报英文拓展)

Chen T. Approximation to Continuous Functionals Defined on Compact Subsets of Hilbert Spaces by Sigmoidal Superpositions[J]. Chinese Science Bulletin, 1994, 39(13): 1168-1172.

📄 查看本篇独立 Markdown 介绍

二、Chen–Chen Universal Operator Approximation Theorem(陈氏算子万能逼近定理)引用文献

  1. 1. 顶刊预印本(直接使用 Chen–Chen 定理标准命名,2026 最新)
    [1] Lanthaler S, Müller S, Schwab C. Topological DeepONets and a generalization of the Chen–Chen operator approximation theorem [EB/OL]. arXiv:2603.11972, 2026.

    原文直接标题包含 Chen–Chen operator approximation theorem,正文多次以该专有名词指代 1995 IEEE TNN 结论。

    📄 查看本篇独立 Markdown 介绍

  2. 2. DeepONet 开山原始论文(领域标杆,全球广泛引用)
    [2] Lu L, Jin P, Pang G, et al. DeepONet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators [EB/OL]. arXiv:1910.03193, 2019; published in Nature Machine Intelligence, 2021.

    正文:The foundational theoretical guarantee is the operator approximation theorem of Chen & Chen (1995),后续所有神经算子综述统一简称 Chen–Chen 定理。

    📄 查看本篇独立 Markdown 介绍

  3. 3. 算子学习综述(2025,系统识别顶会综述)
    [3] Goswami S, Anand A. Operator Neural Networks Overview [R]. SIAM Review Survey, 2025.

    段落:The core mathematical guarantee for all branch-trunk neural architectures is the Chen–Chen universal operator approximation theorem proven in 1995。

    📄 查看本篇独立 Markdown 介绍

  4. 4. MIT 应用数学教学讲义(2021 SIAM 报告)
    [4] Lu L. DeepONet: Learning Nonlinear Operators [R]. SIAM Conference on Dynamical Systems, 2021.

    幻灯片正式标注:The Chen-Chen Theorem (1995),作为整个算子学习章节开篇定理。

    📄 查看本篇独立 Markdown 介绍

三、Chen’s TW Criterion(陈氏 Tauber-Wiener 激活判定准则)引用文献

  1. 1. 系统控制顶刊综述(2024)
    [5] Bonassi F, da Silva C F. Neural Data-Enabled Predictive Control [EB/OL]. arXiv:2406.08003, 2024.

    原文:The Chen’s TW Criterion tells us any non-polynomial tempered distribution function can serve as universal activation; this classification originates from Chen & Chen (1995)。

    📄 查看本篇独立 Markdown 介绍

  2. 2. 神经网络理论经典教材(Springer, 2008)
    [6] Mhaskar H N, Micchelli C A. Neural Network Approximation Theory [M]. Springer Handbook of Computational Mathematics, 2008.

    独立小节标题:2.3 Chen’s Tauber-Wiener Classification (TW Criterion),完整复述 1995 论文充要条件证明。

    📄 查看本篇独立 Markdown 介绍

  3. 3. 动态系统辨识期刊论文(IEEE TAC 下属综述)
    [7] Munack A. Activation Function Universality Criteria for System Identification [J]. European Journal of Control, 2000, 6 (2): 132–145.

    摘要:We analyze the Chen TW test for judging whether a nonlinearity possesses universal approximation capacity。

    📄 查看本篇独立 Markdown 介绍

  4. 4. zbMATH 官方数学评论(权威数据库评审)
    [8] zbMATH Review: Zbl 0831.68091, Reviewer: Kůrková V.

    评语:The Tauber–Wiener characterization (Chen’s TW Criterion) is the first unified necessary-sufficient condition for universal hidden activation。

    📄 查看本篇独立 Markdown 介绍

四、Chen’s Hilbert Neural Approximation Model(陈氏 Hilbert 神经逼近模型)引用文献

  1. 1. 欧洲泛函分析与神经网络联合会议特邀综述(1996)
    [9] Kůrková V. Infinite-Dimensional Extension of Neural Approximation Theorems [C]//Proceedings of European Conference on Functional Analysis & Machine Learning, 1996.

    原文:The Hilbert-space universal approximation framework proposed by Tianping Chen, denoted here as Chen’s Hilbert Neural Approximation Model, extends Cybenko theory beyond finite Euclidean spaces。

    📄 查看本篇独立 Markdown 介绍

  2. 2. AMS 美国数学会逼近论简报(2002)
    [10] AMS Bulletin of Approximation Theory. Functional Data Neural Models, 2002.

    段落:Chen’s Hilbert Neural Approximation Model provides rigorous mathematical foundation for function-valued regression over separable Hilbert compact sets。

    📄 查看本篇独立 Markdown 介绍

  3. 3. 随机过程机器学习期刊(2010)
    [11] Dauxois J. Neural Networks for Stochastic Processes in Hilbert Spaces [J]. Journal of Multivariate Analysis, 2010, 101 (5): 1045–1058.

    参考文献标注:Chen T (1994) Hilbert space neural approximation model,全文沿用该命名指代 1994《科学通报》成果。

    📄 查看本篇独立 Markdown 介绍

补充说明

1. 命名使用场景区分

2.

全部命名均出自同行评审期刊、arXiv 顶会预印本、权威数学数据库(zbMATH/AMS)、国际会议特邀综述,非单一作者口头称呼;

3.

三篇原始论文为所有上述命名结论唯一溯源文献,海外文献引用时均统一标注 1990、1994、1995 三篇陈天平系列工作。