Character Recognition Analysis with Nonnegative Matrix Factorization

AUTHOR(S):

TITLE

Character Recognition Analysis with Nonnegative Matrix Factorization

ABSTRACT

In this paper, we analyze character recognition performance of three different nonnegative matrix factorization (NMF) algorithms. These are multiplicative update (MU) rule known as standard NMF, alternating least square (NMF-ALS) and projected gradient descent (NMF-PGD). They are most preferred approaches in the literature. There are lots of application areas for NMF such as robotics, bioinformatics, vision, sound and others. We use well known MNIST digit data set to test the performance of NMF, NMF-ALS and NMF-PGD. Experimental results show that NMF-ALS is the best and the worst one is NMF-PGD for these there algorithms in the meaning of accuracy. Therefore, we suggest NMF-ALS method can be used to analyze patterns on character recognition.

KEYWORDS

Nonnegative matrix factorization, character recognition, pattern recognition, multiplicative update rule, alternating least square, projected gradient descent

REFERENCES

[1] D. D. Lee and S. Seung, Learning the Parts of Objects by Nonnegative Matrix Factorization, Nature 401, 1999, pp. 788-791.

[2] D. D. Lee and S. Seung, Algorithms for Nonnegative Matrix Factorization, International Conference on Neural Information Processing, 2000.

[3] S. Choi, “Algorithms for orthogonal nonnegative matrix factorization”, in Proc. Int. Joint Conf. on Neural Networks, Hong Kong, Jun. 2008, pp. 1828-1832.

[4] F. Shahnaz, M. W. Berry, V. P. Pauca, and R. J. Plemmons, “Document clustering using nonnegative matrix factorization, Int. Journal of Information Processing and Management, vol. 42, Issue 2, Mar. 2006, pp. 373-386.

[5] T. Ensari, J. Chorowski, and J. M. Zurada, “Correntropy-based document clustering via nonnegative matrix factorization”, in Proc. Int. Conf. on Artificial Neural Networks, vol. 7553, Sept. 2012, pp. 347-354.

[6] X. Li, J. Zhou, L. Tong, X. Yu, J. Guo, C. Zhao, Structured Discriminative Nonnegative Matrix Factorization for Hyperspectral Unmixing, IEEE International Conference on Image Processing, 2016, pp. 1848-1852.

[7] W. Zhao, H. Ma, and N. Li, “A new nonnegative matrix factorization algorithm with sparseness constraints”, in Proc. the Int. Conf. on Machine Learning and Cybernetics, Jul. 2011, pp 1449-1452.

[8] H. Liu, Z. Wu, X. Li, D. Cai, and T. S. Huang, “Constrained nonnegative matrix factorization for image representation”, IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 34, no. 7, Jul. 2012.

[9] C. Fevotte and J. Idier, “Algorithms for nonnegative matrix factorization with the β- Divergence”, Neural Computation, vol. 23, no. 9, Sept. 2011, pp. 2421-2456.

[10] M. Berry, M. Browne, A. M. Langville, V. P. Pauca, R. J. Plemmons, “Algorithms and Applications for Approximate Nonnegative Matrix Factorization”, Int. Journal of Computational Statistics for Approximate Nonnegative Matrix Factorization, vol. 52, no. 1, Sept. 2007, pp. 155-173.

[11] C. J. Lin, “Projected Gradient Methods for Nonnegative Matrix Factorization”, Neural Computation, Neural Computation, vol. 19, no. 10, Aug. 2007, pp. 2756-2779.

[12] H. Kim, H. Park, “Nonnegative Matrix Factorization Based on Alternating Nonnegativity Constrained Least Squares and Active Set Method”, SIAM Journal of Matrix Analysis and Applications, vol. 30, no. 2, 2008, pp. 713-730.

[13] H. Liu, Z. Wu, X. Li, D. Cai, and T. S. Huang, “Constrained nonnegative matrix factorization for image representation”, IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 34, no. 7, Jul. 2012.

[14] D. Guillamet, J. Vitria, and B. Schiele, “Introducing a weighted non-negative matrix factorization for image classification”, Pattern Recognition Letters, vol. 24, pp. 2447-2454, Oct. 2003.

[15] R. He, W. S. Zheng, and B. G. Hu, “Maximum correntropy criterion for robust face recognition”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 8, pp. 1561-1576, Aug. 2011.

[16] R. He, B. G. Hu, W. S. Zheng, and X. W. Kong, “Robust principal component analysis based on maximum correntropy criterion”, IEEE Trans. on Image Processing, vol. 20, no. 6, pp. 1485-1494, Jun. 2011.

[17] R. Chalasani and J. H. Principe, “Self organizing maps with correntropy induced metric”, in Proc. Int. Joint Conf. on Neural Networks, Jul. 2010, pp. 1-6.

[18] W. Zhao, H. Ma, and N. Li, “A new nonnegative matrix factorization algorithm with sparseness constraints”, in Proc. the Int. Conf. on Machine Learning and Cybernetics, Jul. 2011, pp 1449-1452.

[19] P. Tan, M. Steinbach, and V. Kumar, Introduction to Data Mining, Pearson Addison Wesley, May 2005.

[20] M. W. Berry, N. Gillis, and F. Glineur, “Document classification using nonnegative matrix factorization and underapproximation”, in Proc. Int. Symp on Circuits and Systems, May 2009, pp. 2782-2785.

[21] B. J. Shastri and M. D. Levine, “Face recognition using localized features based on non-negative sparse coding”, Machine Vision and Applications, vol. 18, no. 2, pp. 107-122, Apr. 2007.

[22] H. J. Oh, K. M. Lee, and S. U. Lee, “Occlusion invariant face recognition using selective local non-negative matrix factorization basis images”, Image and Vision Computing, vol. 26, Issue 11, pp. 1515-1523, Nov. 2008.

[23] J. Y. Pan and J. S. Zhang, “Large margin based nonnegative matrix factorization and partial least squares regression for face recognition”, Pattern Recognition Letters, vol. 32, pp. 1822- 1835, Oct. 2011.

[24] P. Paatero, “Least squares formulation of robust non-negative factor analysis”, Chemometrics and Intelligent Laboratory Systems 37, pp. 23-35, May 1997.

Cite this paper

Tolga Ensari. (2016) Character Recognition Analysis with Nonnegative Matrix Factorization. International Journal of Computers, 1, 219-222