Concepedia

Gradual Pruning

Gradual Pruning is a model compression technique that systematically reduces a neural network's connectivity over multiple training phases, distinguished from one-shot or abrupt pruning methods by incrementally removing weights according to a predefined schedule or feedback signal.

13

Publications

1K

Citations

50

Authors

18

Institutions

Publications per year

2017–2026

12

Parents

Publications

13

To prune, or not to prune: exploring the efficacy of pruning for model compression

Michael Zhu, Suyog Gupta · arXiv (Cornell University) · 2017 · 663 citations · Full text

, , +16

To prune, or not to prune: exploring the efficacy of pruning for model compression

Michael Zhu, Suyog Gupta · arXiv (Cornell University) · 2017 · 115 citations · Full text

, , +16

WoodFisher: Efficient Second-Order Approximation for Neural Network Compression

Sidak Pal Singh, Dan Alistarh · arXiv (Cornell University) · 2020 · 37 citations · Full text

, , +17

Comprehensive Graph Gradual Pruning for Sparse Training in Graph Neural Networks

Chuang Liu, Xueqi Ma, Yibing Zhan et al. · IEEE Transactions on Neural Networks and Learning Systems · 2023 · 36 citations

, , +15

Gradual Channel Pruning While Training Using Feature Relevance Scores for Convolutional Neural Networks

Sai Aparna Aketi, Sourjya Roy, Anand Raghunathan et al. · IEEE Access · 2020 · 33 citations · Full text

, , +15

1–5 of 13

Authors

50

  1. MZ
    Michael Zhu

    Stanford University, United States

    2 publications · 778 citations

    2

    h-index

  2. SG
    Suyog Gupta

    IBM Research - Thomas J. Watson Research Center, United States

    2 publications · 778 citations

    2

    h-index

  3. DA
    Dan Alistarh

    Institute of Science and Technology Austria, Austria

    2 publications · 51 citations

    2

    h-index

  4. SP
    Sidak Pal Singh

    ETH Zurich, Switzerland

    2 publications · 51 citations

    2

    h-index

  5. SA
    Sai Aparna Aketi

    Purdue University West Lafayette, United States

    1 publications · 33 citations

    1

    h-index

1–5 of 50

Institutions

18

  1. IBM (United States)

    Armonk, United States

    1 h-index

    1 publications

    115 citations

  2. Alibaba Group (United States)

    San Mateo, United States

    1 h-index

    1 publications

    16 citations

  3. Purdue University West Lafayette

    West Lafayette, United States

    1 h-index

    4 publications

    132 citations

  4. The University of Texas at Austin

    Austin, United States

    1 h-index

    4 publications

    52 citations

  5. National University of Kaohsiung

    Kaohsiung City, Taiwan

    1 h-index

    1 publications

    10 citations

  6. University of Twente

    Enschede, The Netherlands

    1 h-index

    2 publications

    26 citations

  7. Jingdong (China)

    Beijing, China

    1 h-index

    2 publications

    72 citations

  8. Yunnan University

    Kunming, China

    1 h-index

    1 publications

    36 citations

1–8 of 18

Venues

  1. arXiv (Cornell University)

    Repository · 5 publications · 842 citations

    5

    h-index

  2. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 1 publications · 22 citations

    1

    h-index

  3. 1

    h-index

  4. IEEE Access

    1 publications · 33 citations

    1

    h-index

  5. Computers & Electrical Engineering

    1 publications · 13 citations

    1

    h-index

1–5