Share Email Print

Proceedings Paper • new

Data clustering based on label consistent constraint matrix factorization
Author(s): Xianzhong Long
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The schemes based on matrix factorization have been applied in many fields, such as recommender system, image classification, data clustering, face recognition and so on. Among them, non negative matrix factorization (NMF) and concept factorization (CF) are the two most commonly used matrix decomposition techniques. NMF decomposes a matrix into the product of two non negative matrices and CF divides a matrix into the product of three matrices. CF is considered as one variant of NMF. The biggest difference between them is that CF can be executed in a kernel space. The development of supervised learning methods show that label information is critical to enhance the model’s ability. In this paper, NMF and CF based on label consistent constraint methods are presented respectively for data clustering. The corresponding multiplicative update solutions, parameters selection and convergence verification are given. Clustering results on three data sets reveal that our methods outperform the state-of-the-art algorithms in terms of accuracy and normalized mutual information.

Paper Details

Date Published: 9 August 2018
PDF: 12 pages
Proc. SPIE 10806, Tenth International Conference on Digital Image Processing (ICDIP 2018), 108066W (9 August 2018); doi: 10.1117/12.2503137
Show Author Affiliations
Xianzhong Long, Nanjing Univ. of Posts and Telecommunications (China)

Published in SPIE Proceedings Vol. 10806:
Tenth International Conference on Digital Image Processing (ICDIP 2018)
Xudong Jiang; Jenq-Neng Hwang, Editor(s)

© SPIE. Terms of Use
Back to Top