IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Zero-Shot Feature Selection via Transferring Supervised Knowledge

Zero-Shot Feature Selection via Transferring Supervised Knowledge
View Sample PDF
Author(s): Zheng Wang (Department of Computer Science and Technology, University of Science and Technology Beijing, Beijing, China), Qiao Wang (School of Software, Tsinghua University, Beijing, China), Tingzhang Zhao (Department of Computer Science and Technology, University of Science and Technology Beijing, Beijing, China), Chaokun Wang (School of Software, Tsinghua University, Beijing, China)and Xiaojun Ye (School of Software, Tsinghua University, Beijing, China)
Copyright: 2021
Volume: 17
Issue: 2
Pages: 20
Source title: International Journal of Data Warehousing and Mining (IJDWM)
Editor(s)-in-Chief: Eric Pardede (La Trobe University, Australia)and Kiki Adhinugraha (La Trobe University, Australia)
DOI: 10.4018/IJDWM.2021040101

Purchase

View Zero-Shot Feature Selection via Transferring Supervised Knowledge on the publisher's website for pricing and purchasing information.

Abstract

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.

Related Content

Qiliang Zhu, Changsheng Wang, Wenchao Jin, Jianxun Ren, Xueting Yu. © 2024. 17 pages.
Feiqi Liu, Dong Yang, Yuyang Zhang, Chengcai Yang, Jingjing Yang. © 2024. 19 pages.
JianDong He. © 2024. 14 pages.
Qiliang Zhu, Wenhao Ding, Mingsen Xiang, Mengzhen Hu, Ning Zhang. © 2023. 16 pages.
Xiaohui Yang, Hailong Ma, Miao Wang. © 2023. 17 pages.
Man Jiang, Qilong Han, Haitao Zhang, Hexiang Liu. © 2023. 15 pages.
Ge Zhang, Zubin Ning. © 2023. 21 pages.
Body Bottom