Unsupervised feature selection using orthogonal locality preserving projections and bipartite graph matching for face image classification

Document Type : Research Paper

Author

Department of Electrical Engineering, Lorestan University, Khoramabad, Iran

Abstract

Feature selection plays a crucial role in facial image classification by reducing dimensionality and improving robustness to variations in expression, pose, and lighting. However, researchers face challenges when selecting features from high-dimensional, unlabeled data due to the nonlinear manifold structure of facial images. To address this, this paper proposes UFSOLPP, a novel unsupervised feature selection method that consists of three main stages. First, the method employs Orthogonal Locality Preserving Projections (OLPP) for feature extraction, aiming to preserve local data structures and enforce orthogonality without dimensionality reduction. Unlike conventional OLPP, which uses heat kernel to measure similarity, this paper replaces it with cosine distance to better capture angular relationships that are for facial image discrimination. Second, it measures the similarity between the original and orthogonal features using the Pearson correlation distance. Third, it models both feature sets as vertices in a weighted bipartite graph. The edge weights are computed using the Pearson correlation similarity, and the method uses the Hungarian algorithm to compute maximum matching. The method selects the original features involved in the maximum matching as the final subset. This strategy removes noisy, correlated, and redundant features effectively, while preserving interpretability and discriminative power. Experiments demonstrate that UFSOLPP outperforms eight state-of-the-art methods. It achieves 96.00% accuracy and 0.9800 NMI on Jaffe, 68.66% accuracy and 0.7532 NMI on ORL, and 82.33% accuracy and 0.8557 NMI on the high-dimensional Pixraw10P dataset. These results highlight the practical value of UFSOLPP and its ability to handle high-dimensional data efficiently in unsupervised facial image analysis.

Keywords

Main Subjects

References

[1] Zhao, X., Wu, D., Nie, F., Member, S., & Yu, W. (2024). Nonlinear Locality-Preserving Projections With Dynamic Graph Learning. IEEE Transactions on Neural Networks and Learning Systems, PP, 1{12. https://doi.org/10.1109/TNNLS.2024.3408835.
[2] Yao, Y., Meng, H., Gao, Y., Long, Z., & Li, T. (2023). Linear dimensionality reduction method based on topological properties. Information Sciences, 624, 493{511. https://doi.org/10.1016/j.ins.2022.12.098.
[3] Fattahi, M., Moattar, M. H., & Forghani, Y. (2023). Locally alignment-based manifold learning for simultaneous feature selection and extraction in classi cation problems. Knowledge-Based Systems, 259, 110088. https://doi.org/10.1016/j.knosys.2022.110088.
[4] Gerber, S., Tasdizen, T., Fletcher, P. T., Joshi, S., Whitaker, R., & Adni, I. (2010). Manifold modeling for brain population analysis. Medical Image Analysis, 14(5), 643{653. https://doi.org/10.1016/j.media.2010.05.0081.
[5] Ahadzadeh, B., Abdar, M., Safara, F., Khosravi, A., Menhaj, M. B., & Suganthan, P. N. (2023). SFE: A Simple, Fast, and Ecient Feature Selection Algorithm for High-Dimensional Data. IEEE Transactions on Evolutionary Computation, 27(6), 1896{1911. https://doi.org/10.1109/TEVC.2023.3238420.
[6] Qin, Z., Chen, H., Mi, Y., Luo, C., Horng, S., & Li, T. (2024). Multi-label feature selection with adaptive graph learning and label information enhancement. Knowledge-Based Systems, 285(December 2023), 111363. https://doi.org/10.1016/j.knosys.2023.111363.
[7] Lee, K. K. G., Kasim, H., Zhou, W. J., Sirigina, R. P., & Hung, G. G. T. (2023). Feature redundancy assessment framework for subject matter experts. Engineering Applications of Arti cial Intelligence, 117(September 2022), 105456. https://doi.org/10.1016/j.engappai.2022.105456.
[8] Afshar, M., & Use , H. (2022). Optimizing feature selection methods by removing irrelevant features using sparse least squares. Expert Systems with Applications, 200(April), 116928. https://doi.org/10.1016/j.eswa.2022.116928.
[9] Beiranvand, F., Mehrdad, V., & Dowlatshahi, M. B. (2022). Unsupervised feature selection for image classi cation: A bipartite matching-based principal component analysis approach. Knowledge-Based Systems, 250, 109085. https://doi.org/10.1016/j.knosys.2022.109085.
[10] Xiang, L., Chen, H., Yin, T., Horng, S. J., & Li, T. (2024). Unsupervised feature selection based on a bipartite graph and low-redundant regularization. Knowledge-Based Systems, 302(June), 112379. https://doi.org/10.1016/j.knosys.2024.112379.
[11] Paniri, M., Dowlatshahi, M. B., & Nezamabadi-pour, H. (2020). MLACO: A multi-label feature selection algorithm based on ant colony optimization. Knowledge-Based Systems, 192, 105285. https://doi.org/10.1016/j.knosys.2019.105285.
[12] Hashemi, A., Dowlatshahi, M. B., & Nezamabadi-pour, H. (2021). An ecient Pareto-based feature selection algorithm for multi-label classi cation. Information Sciences, 581, 428{447. https://doi.org/10.1016/j.ins.2021.09.052.
[13] Ang, J. C., Mirzal, A., Haron, H., & Hamed, H. N. A. (2016). Supervised, unsupervised, and semi-supervised feature selection: A review on gene selection. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 13(5),
971{989. https://doi.org/10.1109/TCBB.2015.2478454.
[14] Hindawi, M., Elghazel, H., & Benabdeslem, K. (2013). Ecient semi-supervised feature selection by an ensemble approach. Unde ned.
[15] Yao, C., Liu, Y. F., Jiang, B., Hen, J., & Han, J. (2017). LLE Score: A New Filter-Based Unsupervised Feature Selection Method Based on Nonlinear Manifold Embedding and Its Application to Image Recognition. IEEE Transactions on Image Processing, 26(11), 5257{5269. https://doi.org/10.1109/TIP.2017.2733200.
[16] Xiao, Z., Chen, H., Mi, Y., Luo, C., Horng, S. J., & Li, T. (2025). Joint subspace learning and subspace clustering based unsupervised feature selection. Neurocomputing, 635(March), 129885. https://doi.org/10.1016/j.neucom.2025.129885.
[17] Wang, Y., Huang, M., Zhou, L., Che, H., & Jiang, B. (2024). Multi-cluster nonlinear unsupervised feature selection via joint manifold learning and generalized Lasso. 255(February).
[18] Li, D., Chen, H., Mi, Y., & Luo, C. (2024). Unsupervised feature selection via dual space-based low redundancy scores and extended OLSDA. 662(December 2023).
[19] Xiang, L., Chen, H., Yin, T., Horng, S. J., & Li, T. (2024). Unsupervised feature selection based on bipartite graph and low-redundant regularization. Knowledge-Based Systems, 302(February), 112379. https://doi.org/10.1016/j.knosys.2024.112379.
[20] Han, K., Wang, Y., Zhang, C., Li, C., & Xu, C. (2018). Autoencoder Inspired Unsupervised Feature Selection. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, 2018-April, 2941{2945. https://doi.org/10.1109/ICASSP.2018.8462261.
[21] Guo, J., & Zhu, W. (2018). Dependence guided unsupervised feature selection. 32nd AAAI Conference on Arti cial Intelligence, AAAI 2018, 2232{2239. https://doi.org/10.1609/aaai.v32i1.11904.
[22] Selection Algorithms Based on Standard Deviation and Cosine Similarity for Genomic Data Analysis. Frontiers in Genetics, 12. https://doi.org/10.3389/fgene.2021.684100.
[23] Zhu, P., Zuo, W., Zhang, L., Hu, Q., & Shiu, S. C. K. (2015). Unsupervised feature selection by regularized self-representation. Pattern Recognition, 48(2), 438{446. https://doi.org/10.1016/j.patcog.2014.08.006.
[24] Huang, D., Cai, X., & Wang, C. D. (2019). Unsupervised feature selection with multi-subspace randomization and collaboration. Knowledge-Based Systems, 182(July). https://doi.org/10.1016/j.knosys.2019.07.027.
[25] Li, X., Zhang, H., Zhang, R., Liu, Y., & Nie, F. (2019). Generalized Uncorrelated Regression with Adaptive Graph for Unsupervised Feature Selection. IEEE Transactions on Neural Networks and Learning Systems, 30(5), 1587{1595. https://doi.org/10.1109/TNNLS.2018.2868847.
[26] Samareh-Jahani, M., Saberi-Movahed, F., Eftekhari, M., Aghamollaei, G., & Tiwari, P. (2024). Low-Redundant Unsupervised Feature Selection based on Data Structure Learning and Feature Orthogonalization. Expert Systems with Applications, 240(October 2023), 122556. https://doi.org/10.1016/j.eswa.2023.122556.
[27] Xiao, Z., Chen, H., Mi, Y., Luo, C., Horng, S. J., & Li, T. (2025). Joint subspace learning and subspace clustering based unsupervised feature selection. Neurocomputing, 635(March), 129885. https://doi.org/10.1016/j.neucom.2025.129885.
[28] Ma, Z., Wei, Y., Huang, Y., & Wang, J. (2024). Unsupervised feature selection based on minimum-redundant subspace learning with self-weighted adaptive graph. Digital Signal Processing: A Review Journal, 155(August), 104738. https://doi.org/10.1016/j.dsp.2024.104738.
[29] Zhang, C., Nie, F., Wang, R., & Li, X. (2024). Supervised Feature Selection via Multi-Center and Local Structure Learning. IEEE Transactions on Knowledge and Data Engineering, 36(9), 4930{4942. https://doi.org/10.1109/TKDE.2024.3372657.
[30] Zohrati, L., Abadeh, M. N., & Kazemi, E. (2018). Flexible approach to schedule tasks in cloud-computing environments. IET Software, 12(6), 474{479. https://doi.org/10.1049/iet-sen.2017.0008.
[31] Cai, D., He, X., Han, J., & Member, S. (2006). Orthogonal Laplacianfaces for Face Recognition. 15(11), 3608{3614.
[32] Liu, B. Di, Wang, Y. X., Zhang, Y. J., & Shen, B. (2013). Learning dictionary on manifolds for image classi cation. Pattern Recognition, 46(7), 1879{1890. https://doi.org/10.1016/j.patcog.2012.11.018.
[33] Ding-cheng, F. (n.d.). Detecting Local Manifold Structure for Unsupervised Feature Selection. Acta Automatica Sinica, 40(10), 2253{2261. https://doi.org/10.1016/S1874-1029(14)60362-1.
[34] Yu, J., & Zhang, C. (2020). Manifold regularized stacked autoencoders-based feature learning for fault detection in industrial processes. Journal of Process Control, 92, 119{136. https://doi.org/10.1016/j.jprocont.2020.06.001.
[35] Yousu , M., Babu, R., & Rathinam, A. (2024). Nonlinear dimensionality reduction based visualization of single-cell RNA sequencing data. Journal of Analytical Science and Technology, 15(1), 1{23. https://doi.org/10.1186/s40543-023-00414-0.
[36] Belkin, M., & Niyogi, P. (2002). Laplacian eigenmaps and spectral techniques for embedding and clustering. Advances in Neural Information Processing Systems. https://doi.org/10.7551/mitpress/1120.003.0080.
[37] Kienitz, D., Komendantskaya, E., & Lones, M. (2022). The E ect of Manifold Entanglement and Intrinsic Dimensionality on Learning. Proceedings of the AAAI Conference on Arti cial Intelligence, 36(7), 7160{7167. https://doi.org/10.1609/aaai.v36i7.20676.
[38] Question, S. (2012). Pearson's correlation coecient. 4483(July), 1{2. https://doi.org/10.1136/bmj.e4483.
[39] Armiti, A., & Gertz, M. (n.d.). Geometric Graph Matching and Similarity: A Probabilistic Approach. Categories and Subject Descriptors.
[40] MANUAL, SOLUTION. Introduction to Graph Theory.
[41] Schrijver, A. (2017). A Course in Combinatorial Optimization.
[42] Dwivedi, S. P. (n.d.). Approximate Bipartite Graph Matching. Springer Singapore. https://doi.org/10.1007/978-981-16-8546-0.
[43] Schwartz, J., Steger, A., & Weil, A. (2005). Fast Algorithms for Weighted Bipartite Matching, 476{487.
[44] Rahimi, Z., Taghipour, K., Khadivi, S., & Afhami, N. (2012). Document and sentence alignment in comparable corpora using bipartite graph matching. In 2012 6th International Symposium on Telecommunications (IST), November, 817{821.
https://doi.org/10.1109/ISTEL.2012.6483098.
[45] Zhu, H., Liu, D., Zhang, S., Zhu, Y., & Teng, L. (2016). Solving the Many to Many assignment problem by improving the Kuhn{Munkres algorithm with backtracking. 618, 30{41. https://doi.org/10.1016/j.tcs.2016.01.002.
[46] Beiranvand, F., Mehrdad, V., & Dowlatshahi, M. B. (2025). Unsupervised feature selection based on the two-dimensional principal component analysis and bipartite graph for face image classi cation. Journal of Mahani Mathematical Research, February 1.
[47] Tan, H., Zhang, X., Guan, N., Tao, D., & Huang, X. (2015). Two-Dimensional Euler PCA, 548{559.
[48] Nene, S. A., Nayar, S. K., & Murase, H. (1996). Columbia Object Image Library (COIL-20), Tech. Rep. CUCS-005-96, Dept. Comput. Sci., Columbia Univ., New York, NY,USA.
[49] Arafat, H., Alfeilat, A., Hassanat, A. B. A., Lasassmeh, O., & Tarawneh, A. S. (2019). E ects of Distance Measure Choice on K-Nearest Neighbor Classi er Performance: A Review, 00(00). https://doi.org/10.1089/big.2018.0175.
[50] Liao, Y., & Vemuri, V. R. (2002). Use of k-nearest neighbor classi er for intrusion detection. Computers & Security, 21(5), 439{448.
[51] Jain, D., & Singh, V. (2018). Feature selection and classi cation systems for chronic disease prediction: A review. Egyptian Informatics Journal, 19(3), 179{189. https://doi.org/10.1016/j.eij.2018.03.002. 

Files

History

  • Receive Date: 14 February 2025
  • Revise Date: 06 August 2025
  • Accept Date: 17 September 2025

Share

How to cite

Statistics