Zobrazeno 1 - 10
of 1 195
pro vyhledávání: '"Zhang Shijun"'
Autor:
Fan, Fenglei, Fan, Juntong, Wang, Dayang, Zhang, Jingbo, Dong, Zelin, Zhang, Shijun, Wang, Ge, Zeng, Tieyong
The rapid growth of large models' size has far outpaced that of GPU memory. To bridge this gap, inspired by the succinct relationship between genotype and phenotype, we turn the model compression problem into the issue of parameter representation to
Externí odkaz:
http://arxiv.org/abs/2409.00592
Autor:
Wang, Qianchao, Zhang, Shijun, Zeng, Dong, Xie, Zhaoheng, Guo, Hengtao, Fan, Feng-Lei, Zeng, Tieyong
In this paper, we propose a new super-expressive activation function called the Parametric Elementary Universal Activation Function (PEUAF). We demonstrate the effectiveness of PEUAF through systematic and comprehensive experiments on various industr
Externí odkaz:
http://arxiv.org/abs/2407.09580
In this work, we propose a balanced multi-component and multi-layer neural network (MMNN) structure to approximate functions with complex features with both accuracy and efficiency in terms of degrees of freedom and computation cost. The main idea is
Externí odkaz:
http://arxiv.org/abs/2407.00765
Publikováno v:
Jixie chuandong, Vol 44, Pp 71-80 (2020)
A double closed loop 5-DOF manipulator arm is proposed, which adds a closed loop structure on the basis of((2-UPS)+U)PU configuration to obtain a new topology configuration 2-UPS+((2-UPS)+U)PU. The introduction of the closed loop st
Externí odkaz:
https://doaj.org/article/fceda02cba374179a3437517454ce9c7
Autor:
Tian, Bohao, Zhang, Shijun, Chen, Sirui, Zhang, Yuru, Peng, Kaiping, Zhang, Hongxing, Wang, Dangxiao
Flow, an optimal mental state merging action and awareness, significantly impacts our emotion, performance, and well-being. However, capturing its swift fluctuations on a fine timescale is challenging due to the sparsity of the existing flow detectin
Externí odkaz:
http://arxiv.org/abs/2310.12035
Publikováno v:
Journal of Machine Learning Research, 25(35):1--39, 2024
This paper explores the expressive power of deep neural networks for a diverse range of activation functions. An activation function set $\mathscr{A}$ is defined to encompass the majority of commonly used activation functions, such as $\mathtt{ReLU}$
Externí odkaz:
http://arxiv.org/abs/2307.06555
In this work, a comprehensive numerical study involving analysis and experiments shows why a two-layer neural network has difficulties handling high frequencies in approximation and learning when machine precision and computation cost are important f
Externí odkaz:
http://arxiv.org/abs/2306.17301
Publikováno v:
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:41452-41487, 2023
This paper explores the expressive power of deep neural networks through the framework of function compositions. We demonstrate that the repeated compositions of a single fixed-size ReLU network exhibit surprising expressive power, despite the limite
Externí odkaz:
http://arxiv.org/abs/2301.12353
Publikováno v:
PLoS ONE. 9/18/2024, Vol. 19 Issue 9, p1-15. 15p.
Publikováno v:
Advances in Neural Information Processing Systems, 35:5669--5681, 2022
This paper proposes a new neural network architecture by introducing an additional dimension called height beyond width and depth. Neural network architectures with height, width, and depth as hyper-parameters are called three-dimensional architectur
Externí odkaz:
http://arxiv.org/abs/2205.09459