Accuracy controlled iterative method for efficient sigmoid function approximation

被引：13
|
作者：
Chiluveru, S. R. [1 ]
Tripathy, M. [1 ]
Mohapatra, B. [1 ]
机构：
[1] Indian Inst Technol Roorkee, Dept Elect Engn, Roorkee, Uttar Pradesh, India
来源：
ELECTRONICS LETTERS | 2020年 / 56卷 / 18期
关键词：
iterative methods; function approximation; piecewise linear techniques; neural nets; optimisation; hardware implementation; daunting job; piecewise linear approximation; linear functions; hardware complexity; accuracy controlled iterative method; sigmoid function approximation; sigmoid activation function; neural networks; NONLINEAR FUNCTION; NEURAL-NETWORKS; LOW-ERROR; HARDWARE; IMPLEMENTATION;
D O I：
10.1049/el.2020.0854
中图分类号：
TM [电工技术]; TN [电子技术、通信技术];
学科分类号：
0808 ; 0809 ;
摘要：
The sigmoid activation function plays an essential role in implementing neural networks in hardware. However, due to a high degree of non-linearity, the hardware implementation of the activation function itself is a daunting job. In this Letter, the authors proposed an accuracy controlled iterative algorithm for the piecewise linear approximation of the sigmoid function. The proposed algorithm approximates sigmoid activation with an optimised number of linear functions for a given error. The accuracy and hardware complexity of the resulting approximation function are compared with the state-of-the-art literature.
引用
收藏
页码：914 / 916
页数：3
相关论文
共 50 条
  • [1] Controlled accuracy approximation of sigmoid function for efficient FPGA-based implementation of artificial neurons
    del Campo, I.
    Finker, R.
    Echanobe, J.
    Basterretxea, K.
    [J]. ELECTRONICS LETTERS, 2013, 49 (25) : 1598 - 1600
  • [2] An iterative calculation method of neuron model with sigmoid function
    Chujo, N
    Kuroyanagi, S
    Doki, S
    Okuma, S
    [J]. 2001 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-5: E-SYSTEMS AND E-MAN FOR CYBERNETICS IN CYBERSPACE, 2002, : 1532 - 1537
  • [3] A Controlled Accuracy-Based Recursive Algorithm for Approximation of Sigmoid Activation
    Samba Raju Chiluveru
    Manoj Tripathy
    Snehit Chunarkar
    [J]. National Academy Science Letters, 2021, 44 : 541 - 544
  • [4] A Controlled Accuracy-Based Recursive Algorithm for Approximation of Sigmoid Activation
    Chiluveru, Samba Raju
    Tripathy, Manoj
    Chunarkar, Snehit
    [J]. NATIONAL ACADEMY SCIENCE LETTERS-INDIA, 2021, 44 (06): : 541 - 544
  • [5] A Novel Approximation Methodology and Its Efficient VLSI Implementation for the Sigmoid Function
    Qin, Zidi
    Qiu, Yuou
    Sun, Huaqing
    Lu, Zhonghai
    Wang, Zhongfeng
    Shen, Qinghong
    Pan, Hongbing
    [J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS, 2020, 67 (12) : 3422 - 3426
  • [6] An Efficient Topology Description Function Method Based on Modified Sigmoid Function
    Yang, Xingfa
    Liu, Jie
    Yang, Yin
    Qing, Qixiang
    Wen, Guilin
    [J]. MATHEMATICAL PROBLEMS IN ENGINEERING, 2018, 2018
  • [7] A Modular Approximation Methodology for Efficient Fixed-Point Hardware Implementation of the Sigmoid Function
    Pan, Zhe
    Gu, Zonghua
    Jiang, Xiaohong
    Zhu, Guoquan
    Ma, De
    [J]. IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, 2022, 69 (10) : 10694 - 10703
  • [8] Efficient iterative solution of electromagnetic scattering using adaptive cross approximation enhanced characteristic basis function method
    Chen, Xinlei
    Gu, Changqing
    Li, Zhuo
    Niu, Zhenyi
    [J]. IET MICROWAVES ANTENNAS & PROPAGATION, 2015, 9 (03) : 217 - 223
  • [9] Approximation of sigmoid function and the derivative for artificial neurons
    Basterrextea, Koldo
    Tarela, José Manuel
    Campo, Inés Del
    [J]. Advances in Neural Networks and Applications, 2001, : 397 - 401
  • [10] On the approximation of the step function by some sigmoid functions
    Iliev, A.
    Kyurkchiev, N.
    Markov, S.
    [J]. MATHEMATICS AND COMPUTERS IN SIMULATION, 2017, 133 : 223 - 234
12345