Zero-Shot Feature Selection via Transferring Supervised Knowledge

被引:0
|
作者
Wang, Zheng [1 ]
Wang, Qiao [2 ]
Zhao, Tingzhang [1 ]
Wang, Chaokun [2 ]
Ye, Xiaojun [2 ]
机构
[1] Univ Sci & Technol Beijing, Dept Comp Sci & Technol, Beijing, Peoples R China
[2] Tsinghua Univ, Sch Software, Beijing, Peoples R China
基金
中国国家自然科学基金; 国家重点研发计划; 中国博士后科学基金;
关键词
Data Mining; Feature Selection;
D O I
10.4018/IJDWM.2021040101
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Feature selection, an effective technique for dimensionality reduction, plays an important role in many machine learning systems. Supervised knowledge can significantly improve the performance. However, faced with the rapid growth of newly emerging concepts, existing supervised methods might easily suffer from the scarcity and validity of labeled data for training. In this paper, the authors study the problem of zero-shot feature selection (i.e., building a feature selection model that generalizes well to "unseen" concepts with limited training data of "seen" concepts). Specifically, they adopt class-semantic descriptions (i.e., attributes) as supervision for feature selection, so as to utilize the supervised knowledge transferred from the seen concepts. For more reliable discriminative features, they further propose the center-characteristic loss which encourages the selected features to capture the central characteristics of seen concepts. Extensive experiments conducted on various real-world datasets demonstrate the effectiveness of the method.
引用
收藏
页码:1 / 20
页数:20
相关论文
共 50 条
  • [21] Transductive zero-shot image classification based on self-supervised enhancement feature
    Wang H.-Y.
    Zhang X.-R.
    Wang X.-S.
    Cheng Y.-H.
    Kongzhi yu Juece/Control and Decision, 2024, 39 (05): : 1707 - 1717
  • [22] Class knowledge overlay to visual feature learning for zero-shot image classification
    Xie, Cheng
    Zeng, Ting
    Xiang, Hongxin
    Li, Keqin
    Yang, Yun
    Liu, Qing
    COMPUTER VISION AND IMAGE UNDERSTANDING, 2021, 207
  • [23] Zero-shot classification based on attribute correlation graph regularized feature selection
    School of Information and Electrical Engineering, China University of Mining & Technology, Xuzhou
    Jiangsu
    221116, China
    不详
    Jiangsu
    221004, China
    Zhongguo Kuangye Daxue Xuebao, 6 (1097-1104):
  • [24] Zero-Shot Learning via Contrastive Learning on Dual Knowledge Graphs
    Wang, Jin
    Jiang, Bo
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS (ICCVW 2021), 2021, : 885 - 892
  • [25] Enhancing Zero-Shot Stance Detection via Targeted Background Knowledge
    Zhu, Qinglin
    Liang, Bin
    Sun, Jingyi
    Du, Jiachen
    Zhou, Lanjun
    Xu, Ruifeng
    PROCEEDINGS OF THE 45TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '22), 2022, : 2070 - 2075
  • [26] Feature Generating Networks for Zero-Shot Learning
    Xian, Yongqin
    Lorenz, Tobias
    Schiele, Bernt
    Akata, Zeynep
    2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2018, : 5542 - 5551
  • [27] Feature Selection via Transferring Knowledge Across Different Classes
    Wang, Zheng
    Ye, Xiaojun
    Wang, Chaokun
    Yu, Philip S.
    ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA, 2019, 13 (02)
  • [28] Zero-shot semi-supervised learning for pansharpening
    Cao, Qi
    Deng, Liang-Jian
    Wang, Wu
    Hou, Junming
    Vivone, Gemine
    INFORMATION FUSION, 2024, 101
  • [29] A Semantic Similarity Supervised Autoencoder for Zero-Shot Learning
    Shen, Fengli
    Lu, Zhe-Ming
    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, 2020, E103D (06): : 1419 - 1422
  • [30] Zero-shot image classification via Visual-Semantic Feature Decoupling
    Sun, Xin
    Tian, Yu
    Li, Haojie
    MULTIMEDIA SYSTEMS, 2024, 30 (02)