Abstract
Neighborhood graph construction plays a vital role in the success of the manifold learning. This paper introduces a supervised hierarchical neighborhood graph construction method for manifold learning. The goal of our method is to discriminate among profoundly close data points with different classes. Our proposed method includes two main phases. The first phase consists of a supervised form of neighborhood selection. The first phase aims to incorporate the Euclidean distance between two objects and their class labels for providing a reasonable estimate of the data’s local topology. The second phase optimizes the neighborhood graph based on the proposed hierarchical search method to find transitive similarity between objects. The second phase aims to improve the representation of the global topology of the manifold. Our proposed method defines a function that compares two data points in a semantically relevant way. The proposed method helps to identify the right neighborhood of each object, and thus, the resulting graph best approximates the actual structure of the manifold. In the experiments, we used seven state-of-the-art works on challenging synthetic and real-world data sets to demonstrate the superiority of the proposed method. In face recognition, our neighborhood graph can improve the canonical LLE and Isomap more than 20 and 50%, respectively. Also, LLE and Isomap based on our proposed method achieve more than 24 and 16% better results respect to the best result of state-of-the-art graph-based manifold learning methods.
Similar content being viewed by others
Notes
Available at: https://lvdmaaten.github.io/drtoolbox/.
References
Zheng, H., Horace, H.S.Ip: Image classification and annotation based on robust regularized coding. SIVP 10(1), 55–64 (2016)
Wei, Z., Chu, Y., Zhao, L.: An effective two-dimensional linear discriminant analysis with locality preserving approach for image recognition. SIVP 11(8), 1577–1584 (2017)
Yildiz, K.: Dimensionality reduction-based feature extraction and classification on fleece fabric images. SIVP 11(2), 317–323 (2016)
Vretos, N., Tefas, A., Pitas, I.: A novel dimensionality reduction technique based on kernel optimization through graph embedding. SIVP 9(1), 3–10 (2015)
Dahmane, A., Larabi, S., Bilasco, L.M., Djeraba, C.: Head pose estimation based on face symmetry analysis. SIVP 9(8), 1871–1880 (2015)
Lu, J., Tan, Y.-P., Wang, G.: Discriminative multimanifold analysis for face recognition from a single training sample per person. IEEE Trans. Pattern Anal. Mach. Intell. 35(1), 39–51 (2013)
Campanella, M., Leonardi, R., Migliorati, P.: Interactive visualization of video content and associated description for semantic annotation. SVIP 3(2), 183–196 (2009)
Berka, T., Vajteršic, M.: Parallel rare term vector replacement: fast and effective dimensionality reduction for text. J. Parallel Distrib. Comput. 73, 341–351 (2013)
Maaten, L.V.D., Postma, E.: Dimensionality Reduction: A Comparative Review. TiCC, Tilburg University, Tilburg (2009)
Duda, R., Hart, P., Stork, D.: Pattern Classification. Wiley, New York (2000)
Tenenbaum, J.B., Silva, Vd, Langford, J.C.: A global geometric framework for nonlinear dimensionality reduction. Science 290(5500), 2319–2323 (2000)
Belkin, M., Niyogi, P.: Laplacian eigenmaps for dimensional reduction and data representation. Neural Comput. 15, 1373–1396 (2000)
Axler, S., Ribet, K.A.: An Introduction to Manifolds, 2nd edn. Universitext: Springer, Berlin (2008)
Roweis, S.T., Saul, L.K.: Nonlinear dimensionality reduction by locally linear embedding. Science 290, 2323–2326 (2000)
Aeini, F., Eftekhari-Moghadam, A.M., Mahmoudi, F.: Non linear dimensional reduction method based on supervised neighborhood graph. In: 7th International Symposium on Telecommunications (IST’2014), IEEE, pp. 35–40 (2014)
Cheng, J., Liu, Q., Lu, H., Chen, Y.W.: Supervised kernel locality preserving projections for face recognition. Neurocomputing 67, 443–449 (2005)
Geng, X., Zhan, D.C., Zhou, Z.H.: Supervised nonlinear dimensionality reduction for visualization and classification. IEEE Trans. Syst. Man Cybern. Part B Cybern. 35(6), 1098–1107 (2005)
Raducanu, B., Dornaika, F.: A supervised non-linear dimensionality reduction approach for manifold learning. Pattern Recognit. 45, 2432–2444 (2012)
Jolliffe, I.T.: Principal Component Analysis, 2nd edn. Springer, New York (2002)
Fukunnaga, K.: Introduction to Statistical Pattern Recognition, 2nd edn. Academic Press, San Diego (1990)
Cox, T.F., Hart, P.E.: Multidimensional Scaling. Chapman and Hall, London (1994)
Silva, D., Tenenbaum, J.B.: Global versus local methods in nonlinear dimensionality reduction, In: Advances in Nerual Information Processing Systems (NIPS), pp. 705–712 (2002)
Weinberger, K., Saul, L.K.: Unsupervised learning of image manifolds by semidefinite programming. In: IEEE International Conference on Computer Vision and Pattern Recognition, pp. 988–995 (2004)
Saul, L.K., Roweis, S.T.: Think globally, fit locally: unsupervised learning of low dimensional manifold. J. Mach. Learn. Res. 4, 119–155 (2003)
Wen, G., Jiang, L., Shadbolt, N.R.: Using graph algebra to optimize neighborhood for isometric mapping. In: IJCAI, pp. 2398–2403 (2007)
Vlachos, M., Domeniconi, M., Gunopulos, D., Kollios, G., Koudas, N.: Non-linear dimensionality reduction techniques for classification and visualization. In: Proceedings of the 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Edmonton, AB, Canada, pp. 645–651 (2002)
Ridder, D.D., Kouropteva, O., Okun, O., Pietikainen, M., Duin, R.P.W.: Supervised locally linear embedding. In: Artificial Neural Networks and Neural Information Processing-ICANN/ICONIP 2003, pp. 333–341, Springer (2003)
Xian-Fa, C., Gui-Hua, W., Jia, W., Jie, L.: Enhanced supervised locality preserving projections for face recognition. In: International Conference on Machine Learning and Cybernetics, IEEE (2011). https://doi.org/10.1109/ICMLC.2011.6017017
Ptucha, R., Savakis, A.: Manifold based sparse representation for facial understanding in natural images. Image Vision Comput. 31, 365–378 (2013)
Schölkopf, B., Smola, A., Müller, K.-R.: Kernel principal component analysis. In: Artificial Neural Networks-ICANN’97, pp. 583–588. Springer, Lausanne (1997)
Hettiarachchi, R., Peters, J.F.: Multi-manifold LLE learning in pattern recognition. Pattern Recognit. 48(9), 2947–2960 (2015)
Omranpour, H., Ghidary, S.S.: Manifold based map representation for mobile robot using Euclidean data difference dimension reduction. Eng. Appl. Artif. Intell. 45, 234–245 (2015)
Lee, C.-S., Elgammal, A., Torki, M.: Learning representations from multiple manifolds. Pattern Recognit. 50, 74–87 (2016)
Nane, S.A., Nayar, S.K., Murase, H.: Columbia object image library (coil-20). Technical Report CUCS-005-96 (1996)
Gao, W., et al.: The CAS-PEAL large-scale Chinese face database and baseline evaluations. IEEE Trans. Syst. Man Cybern. Part A Syst. Hum. 38(1), 149–161 (2008)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Aeini, F., Eftekhari Moghadam, A.M. & Mahmoudi, F. Supervised hierarchical neighborhood graph construction for manifold learning. SIViP 12, 799–807 (2018). https://doi.org/10.1007/s11760-017-1222-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11760-017-1222-4