Publications
[number]
(Position & Institution, Date of first submission) Author(s). Paper title. Journal or conference reference. [ links ]
* Corresponding author; † Equal contribution
NUS = National University of Singapore;
RF = Research Fellow; Duke = Duke University; ARP = Assistant Research Professor;
PolyU = The Hong Kong Polytechnic University; AP = Assistant Professor (Tenure-Track)
Preprints
-
[14]
(AP at PolyU, 12 Jul 2024)
Qianchao Wang†, Shijun Zhang†, Dong Zeng, Zhaoheng
Xie, Hengtao Guo, Feng-Lei Fan, Tieyong Zeng.
Don’t Fear Peculiar Activation Functions: EUAF and Beyond.
Submitted.
[
PDF
arXiv
BibTex
]
-
[13]
(ARP at Duke, 30 Jun 2024)
Shijun Zhang*, Hongkai Zhao, Yimin Zhong, Haomin Zhou.
Structured and Balanced Multi-component and Multi-layer Neural Networks.
Submitted.
[
PDF
arXiv
BibTex
]
-
[12]
(ARP at Duke, 29 Jun 2023)
Shijun Zhang*, Hongkai Zhao, Yimin Zhong, Haomin Zhou.
Why shallow networks struggle with approximating and learning high frequency: A numerical study.
Submitted.
[
PDF
arXiv
BibTex
]
Published (Accepted)
-
[11]
(ARP at Duke, 13 Jul 2023)
Shijun Zhang*, Jianfeng Lu, Hongkai Zhao.
Deep network approximation: Beyond ReLU to diverse activation functions.
Journal of Machine Learning Research, 25(35):1−39, 2024.
[
PDF
arXiv
Journal
BibTex
]
-
[10]
(ARP at Duke, 29 Jan 2023)
Shijun Zhang*, Jianfeng Lu, Hongkai Zhao.
On enhancing expressive power via
compositions of single fixed-size ReLU network.
Proceedings of the 40th International Conference on Machine Learning (ICML 2023), PMLR 202:41452−41487, 2023.
[
PDF
arXiv
Poster
Conference
BibTex
]
-
[9]
(RF at NUS, 19 May 2022)
Zuowei Shen, Haizhao Yang, Shijun Zhang*.
Neural network architecture beyond width and depth.
Advances in Neural Information Processing Systems (NeurIPS 2022), 35:5669−5681, 2022.
[
PDF
arXiv
Poster
Conference
BibTex
]
-
[8]
(RF at NUS, 15 Nov 2021)
Zuowei Shen, Haizhao Yang, Shijun Zhang*.
Deep network approximation in terms of intrinsic parameters.
Proceedings of the 39th International Conference on Machine Learning (ICML 2022), PMLR 162:19909−19934, 2022.
[
PDF
arXiv
Spotlight
Conference
BibTex
]
-
[7]
(RF at NUS, 6 Jul 2021)
Zuowei Shen, Haizhao Yang, Shijun Zhang*.
Deep network approximation:
Achieving arbitrary accuracy with fixed number of neurons.
Journal of Machine Learning Research, Volume 23, Issue 276, September 2022, Pages 1−60.
[
PDF
arXiv
Journal
BibTex
]
-
[6]
(RF at NUS, 28 Feb 2021)
Zuowei Shen, Haizhao Yang, Shijun Zhang*.
Optimal approximation rate of ReLU networks in terms of width and depth.
Journal de Mathématiques Pures et Appliquées, Volume 157, January 2022, Pages 101−135.
[
PDF
arXiv
Journal
BibTex
]
-
[5]
(Ph.D. at NUS, 25 Oct 2020)
Zuowei Shen, Haizhao Yang, Shijun Zhang.
Neural network approximation: Three hidden layers are enough.
Neural Networks, Volume 141, September 2021, Pages 160−173.
[
PDF
arXiv
Journal
BibTex
]
-
[4]
(Ph.D. at NUS, 22 Jun 2020)
Zuowei Shen, Haizhao Yang, Shijun Zhang.
Deep network with approximation error
being reciprocal of width to power of square root of depth.
Neural Computation, Volume 33, Issue 4, April 2021, Pages 1005−1036.
[
PDF
arXiv
Journal
BibTex
]
-
[3]
(Ph.D. at NUS, 9 Jan 2020)
Jianfeng Lu, Zuowei Shen, Haizhao Yang, Shijun Zhang.
Deep network approximation for smooth functions.
SIAM
Journal on Mathematical Analysis, Volume 53, Issue 5, September 2021, Pages 5465−5506.
[
PDF
arXiv
Journal
BibTex
]
-
[2]
(Ph.D. at NUS, 13 Jun 2019)
Zuowei Shen, Haizhao Yang, Shijun Zhang.
Deep network approximation characterized by number of nuerons.
Communications in Computational Physics,
Volume 28, Issue 5, November 2020, Pages 1768−1811.
[
PDF
arXiv
Journal
BibTex
]
-
[1]
(Ph.D. at NUS, 26 Feb 2019)
Zuowei Shen, Haizhao Yang, Shijun Zhang.
Nonlinear approximation via compositions.
Neural Networks,
Volume 119, November 2019, Pages 74−84.
[
PDF
arXiv
Journal
BibTex
]