Deep Active Learning by Leveraging Training Dynamics

被引:0
|
作者
Wang, Haonan [1 ]
Huang, Wei [2 ]
Wu, Ziwei [1 ]
Margenot, Andrew [1 ]
Tong, Hanghang [1 ]
He, Jingrui [1 ]
机构
[1] Univ Illinois, Champaign, IL 61820 USA
[2] Univ New South Wales, Sydney, NSW, Australia
基金
美国国家科学基金会; 美国食品与农业研究所;
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Active learning theories and methods have been extensively studied in classical statistical learning settings. However, deep active learning, i.e., active learning with deep learning models, is usually based on empirical criteria without solid theoretical justification, thus suffering from heavy doubts when some of those fail to provide benefits in real applications. In this paper, by exploring the connection between the generalization performance and the training dynamics, we propose a theory-driven deep active learning method (dynamicAL) which selects samples to maximize training dynamics. In particular, we prove that the convergence speed of training and the generalization performance are positively correlated under the ultra-wide condition and show that maximizing the training dynamics leads to better generalization performance. Furthermore, to scale up to large deep neural networks and data sets, we introduce two relaxations for the subset selection problem and reduce the time complexity from polynomial to constant. Empirical results show that dynamicAL not only outperforms the other baselines consistently but also scales well on large deep learning models. We hope our work would inspire more attempts on bridging the theoretical findings of deep networks and practical impacts of deep active learning in real applications.
引用
收藏
页数:14
相关论文
共 50 条
  • [21] Leveraging Deep Learning for Computer Vision: A Review
    Alam, Ekram
    Abu Sufian
    Das, Akhil Kumar
    Bhattacharya, Arijit
    Ali, Md Firoj
    Rahman, M. M. Hafizur
    [J]. 2021 22ND INTERNATIONAL ARAB CONFERENCE ON INFORMATION TECHNOLOGY (ACIT), 2021, : 298 - 305
  • [22] Leveraging deep contrastive learning for semantic interaction
    Belcaid M.
    Martinez A.G.
    Leigh J.
    [J]. PeerJ Computer Science, 2022, 8
  • [23] Leveraging deep contrastive learning for semantic interaction
    Belcaid, Mahdi
    Martinez, Alberto Gonzalez
    Leigh, Jason
    [J]. PEERJ COMPUTER SCIENCE, 2022, 8
  • [24] Leveraging Deep Learning for Network Anomaly Detection
    Kourtis, Michail-Alexandros
    Oikonomakis, Andreas
    Papadopoulos, Dimitris
    Xylouris, George
    Chochliouros, Ioannis P.
    [J]. 2021 SIXTH INTERNATIONAL CONFERENCE ON FOG AND MOBILE EDGE COMPUTING (FMEC), 2021, : 91 - 96
  • [25] Leveraging Deep Learning for IoT Transceiver Identification
    Gao, Jiayao
    Fan, Hongfei
    Zhao, Yumei
    Shi, Yang
    [J]. ENTROPY, 2023, 25 (08)
  • [26] Leveraging Active Learning for Failure Mode Acquisition
    Kulkarni, Amol
    Terpenny, Janis
    Prabhu, Vittaldas
    [J]. SENSORS, 2023, 23 (05)
  • [27] Phishing Detection Leveraging Machine Learning and Deep Learning: A Review
    Divakaran, Dinil Mon
    Oest, Adam
    [J]. IEEE SECURITY & PRIVACY, 2022, 20 (05) : 86 - 95
  • [28] Virtual learning environment to predict withdrawal by leveraging deep learning
    Saeed-Ul Hassan
    Waheed, Hajra
    Aljohani, Naif R.
    Ali, Mohsen
    Ventura, Sebastian
    Herrera, Francisco
    [J]. INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS, 2019, 34 (08) : 1935 - 1952
  • [29] Pre-Training Acquisition Functions by Deep Reinforcement Learning for Fixed Budget Active Learning
    Yusuke Taguchi
    Hideitsu Hino
    Keisuke Kameyama
    [J]. Neural Processing Letters, 2021, 53 : 1945 - 1962
  • [30] Pre-Training Acquisition Functions by Deep Reinforcement Learning for Fixed Budget Active Learning
    Taguchi, Yusuke
    Hino, Hideitsu
    Kameyama, Keisuke
    [J]. NEURAL PROCESSING LETTERS, 2021, 53 (03) : 1945 - 1962