Positive and Negative Label-Driven Nonnegative Matrix Factorization

Wenhui WU, Yuheng JIA, Shiqi WANG, Ran WANG, Hongfei FAN, Sam KWONG

Research output: Journal PublicationsJournal Article (refereed)peer-review

17 Citations (Scopus)

Abstract

Positive label is often used as the supervisory information in the learning scenario, which refers to the category that a sample is assigned to. However, another side information lying in the labels, which describes the categories that a sample is exclusive of, have been largely ignored. In this paper, we propose a nonnegative matrix factorization (NMF) based classification method leveraging both positive and negative label information, which is termed as positive and negative label-driven NMF (PNLD-NMF). The proposed scheme concurrently accomplishes data representation and classification in a joint manner. Owing to the complementary characteristics between positive and negative labels, we further design a new regularization framework to take advantage of these two label types. Extensive experiments on six image classification benchmark datasets show that the proposed scheme is able to consistently deliver better classification accuracy.
Original languageEnglish
Pages (from-to)2698-2710
JournalIEEE Transactions on Circuits and Systems for Video Technology
Volume31
Issue number7
Early online date29 Sept 2020
DOIs
Publication statusPublished - Jul 2021
Externally publishedYes

Bibliographical note

This work was supported in part by the National Natural Science Foundation of China under Grant 62006158, Grant 61772344, and Grant 61732011; in part by the Hong Kong RGC Early Career Scheme 9048122 under Grant CityU 21211018; in part by the General Research Funds under Grant 9042957 and Grant CityU 11203220; and in part by the Interdisciplinary Innovation Team of Shenzhen University.

Keywords

  • classification
  • negative label
  • Semi-supervised nonnegative matrix factorization

Fingerprint

Dive into the research topics of 'Positive and Negative Label-Driven Nonnegative Matrix Factorization'. Together they form a unique fingerprint.

Cite this