Unsupervised feature selection via Diversity-induced Self-representation
Introduction
The high-dimensional data is ubiquitous in many areas, such as computer vision, pattern recognition and data mining. It not only significantly increases the computation and storage cost, but also induces overfitting and incomprehensible models. To overcome these issues, feature selection techniques have been considered as a type of effective method to reduce the dimensionality by removing irrelevant and redundant features. The aim of feature selection is to obtain a subset of features by removing the noise and redundancy in original features, so that the more intrinsic representation of data and the better performance are achieved [1].
According to the availability of the label information, feature selection approaches can be categorized into supervised methods [2], [3], [4], [5] and unsupervised methods [6], [7], [8], [9], [10], [11]. Compared with supervised methods, unsupervised feature selection methods aim at selecting relevant features without label information. Since the labeling of samples is usually expensive, we cannot always obtain the labels beforehand. Thus, unsupervised feature selection holds great potential in real-world applications.
The early unsupervised feature sePh.D.lection algorithms consider feature ranking techniques as the principle criteria for feature selection [8], [12], [13], [14], [15]. One of the main limitations of these methods is that they treat features independently without considering possible correlations among features. To address this problem, a series of algorithms [10], [16], [17] have been developed. A typical method is spectral clustering based algorithms, which can select a feature subset with preserving the underlying structure between clusters. Spectral clustering based methods explore the cluster structure of data using matrix factorization for spectral analysis, and then select features via sparsity regularization models. Nevertheless, they heavily rely on the learned graph Laplacian. Noises in features may lead to their unreliability. Recently, the self-representation technique has shown significant potential in many tasks, such as subspace clustering [18], [19] and active learning [20], [21]. Motivated by this, some researchers consider the feature selection from the perspective of self-representation property of features [22], i.e., each feature can be well represented by the linear combination of its relevant features. However, they mainly consider selecting the representative features while ignoring the diversity among them. Both representativeness and diversity properties are very important for selecting the effective features: (1) the ideal selected features are supposed to represent the whole original features. That is to say, the highly irrelevant features are discarded, and meanwhile the most relevant features are preserved. (2) The ideal selected features should be diverse enough to capture not only important (representative) but also comprehensive information from features. By considering the diversity property, we can capture more information of data, because features usually describe different aspects of data. (3) The diversity property implies that the very similar features should not be selected simultaneously, so that the redundancy can be greatly reduced. Therefore, there is a great need for integrating both representativeness and diversity properties of features for feature selection.
In this paper, considering both representativeness and diversity properties of features, we propose a novel method, called Diversity-induced Self-representation (DISR) for unsupervised feature selection. Specifically, based on self-representation property, i.e., each feature can be well approximated by the linear combination of its relevant features, the most representative features can be selected. Meanwhile, by incorporating the similarities between features to adjust the weights of being selected, we introduce a diversity term to reduce the redundancy. Then, an efficient optimization algorithm is provided by using the inexact Augmented Lagrange Method (ALM).
Finally, we evaluate our method in both clustering and classification tasks. Experimental results on the synthetic dataset and nine real-world datasets show that the proposed DISR has a better performance than other compared methods.
To summarize, the main contributions of this paper are as follows:
- •
A novel Diversity-induced Self-representation (DISR) for unsupervised feature selection algorithm is proposed. The algorithm considers both the representativeness and diversity properties of features, and hence it can select more valuable features.
- •
A diversity term is introduced to the method. This term is the measure of diversity of all the features. Thus, the diversity of features is used to guide for feature selection.
- •
An iterative algorithm based on inexact ALM is proposed to efficiently solve the optimization model. Experimental results demonstrate that the superiority of our algorithm compared with state-of-the-art algorithms.
The rest of the paper is organized as follows. A brief review of the related works on unsupervised feature selection is given in Section 2. Section 3 introduces the proposed DISR algorithm, and Section 4 describes the optimization of our algorithm in details. Extensive experiments on synthetic and real-world datasets are presented in Section 5. Finally, Section 6 concludes this paper.
Section snippets
Related work
Recently, many unsupervised feature selection methods have been proposed. These methods can be roughly divided into three categories: filter, wrapper, and embedded methods. Because our work belongs to the embedded method, we just briefly review the filter and wrapper methods first, and then review the related works on embedded methods in more details. Filter methods use feature ranking techniques as the principle criteria for feature selection due to their simplicity and practical success
Diversity-induced Self-representation for unsupervised feature selection
In this section, we first introduce the unsupervised feature selection based on regularized self-representation, and then provide our novel unsupervised feature selection method via Diversity-induced Self-representation.
Optimization
So far, the method based on the Diversity-induced Self-representation has been proposed. Obviously, it is hard to find the global optimizers since two non-smooth terms are involved. Thus, we employ the inexact augmented Lagrange method [40] to optimize each variable iteratively.
Experiments
In this section, we demonstrate the effectiveness of the proposed DISR on the synthetic dataset, and then evaluate the performance of DISR in both clustering and classification tasks on the real-world datasets.
Conclusion
In this paper, we propose a novel unsupervised feature selection method via Diversity-induced Self-representation, called DISR, which can select the features with both representativeness and diversity. By incorporating the similarities between features to adjust the weights of being selected, a novel diversity term is designed to eliminate redundancy among selected features. To solve the proposed optimization problem, an efficient optimization algorithm is presented by using ALM method.
Yanbei Liu received the B.E. degree from Zhengzhou University of Light Industry, Zhengzhou, China, in 2009 and the M.E. degree from Tianjin Polytechnic University, Tianjin, China, in 2012. He is currently pursuing the Ph.D. degree from the School of Electronic Information Engineering, Tianjin University, Tianjin, China. His current research interests include machine learning, subspace learning, pattern recognition.
References (51)
- et al.
An efficient feature selection algorithm for hybrid data
Neurocomputing
(2016) - et al.
Global mutual information-based feature selection approach using single-objective and multi-objective optimization
Neurocomputing
(2015) - et al.
Udsfs: unsupervised deep sparse feature selection
Neurocomputing
(2016) - et al.
Group sparse feature selection on local learning based clustering
Neurocomputing
(2016) - et al.
Visual search reranking with relevant local discriminant analysis
Neurocomputing
(2016) - et al.
Unsupervised feature selection by regularized self-representation
Pattern Recognit.
(2015) - et al.
On the approximability of minimizing nonzero variables or unsatisfied relations in linear systems
Theor. Comput. Sci.
(1998) - et al.
An introduction to variable and feature selection
J. Mach. Learn. Res.
(2003) - et al.
On similarity preserving feature selection
IEEE Trans. Knowl. Data Eng.
(2013) - F. Nie, H. Huang, X. Cai, C.H. Ding, Efficient and robust feature selection via joint ℓ2,1-norms minimization, in:...
Unsupervised feature selection using feature similarity
IEEE Trans. Pattern Anal. Mach. Intell.
Sparse subspace clusteringalgorithm, theory, and applications
IEEE Trans. Pattern Anal. Mach. Intell.
Selection of variables to preserve multivariate data structure, using principal components
Appl. Stat.
Bayesian feature and model selection for gaussian mixture models
IEEE Trans. Pattern Anal. Mach. Intell.
Cited by (34)
Unsupervised feature selection via dual space-based low redundancy scores and extended OLSDA
2024, Information SciencesUnsupervised feature selection algorithm based on redundancy learning and sparse regression
2023, Physica A: Statistical Mechanics and its ApplicationsSemi-supervised feature selection via adaptive structure learning and constrained graph learning
2022, Knowledge-Based SystemsAdaptive graph learning for semi-supervised feature selection with redundancy minimization
2022, Information SciencesUnsupervised feature selection with robust data reconstruction (UFS-RDR) and outlier detection
2022, Expert Systems with ApplicationsUnsupervised feature selection via self-paced learning and low-redundant regularization
2022, Knowledge-Based SystemsCitation Excerpt :Given the fact that redundancy between features plays an important role in feature selection, it is necessary to introduce a regularization term to eliminate its negative impact. For this reason, Liu et al. proposed a novel term taking the pairwise similarity of features into consideration [23]. In order to achieve low redundancy between features, we add the pairwise similarity as regularizer to the framework of subspace learning.
Yanbei Liu received the B.E. degree from Zhengzhou University of Light Industry, Zhengzhou, China, in 2009 and the M.E. degree from Tianjin Polytechnic University, Tianjin, China, in 2012. He is currently pursuing the Ph.D. degree from the School of Electronic Information Engineering, Tianjin University, Tianjin, China. His current research interests include machine learning, subspace learning, pattern recognition.
Kaihua Liu received the B.E. degree in 1981, M.E. degree in 1991, and Ph.D. degree in 1999 from Tianjin University, Tianjin, China. Currently, he is a Professor at the School of Electronic Information Engineering, Tianjin University, Tianjin, China. His current research interests include radio frequency identification theory and application, digital signal processing theory and application, pattern recognition.
Changqing Zhang received the B.S. and M.E. degrees in computer science from Sichuan University in 2005 and 2008, and the Ph.D. degree from Tianjin University in 2016, respectively. He is currently an Assistant Professor with Tianjin University. His current research interests include machine learning, data mining, and computer vision.
Jing Wang is currently pursuing the Ph.D. degree from the Faculty of Science and Technology, Bournemouth University, UK. Before that, she received the M.E. degree from City University of Hong Kong, China. Her current research interests include machine learning and data mining, such as nonnegative matrix factorization, subspace clustering and semi-supervised learning.
Xiao Wang received M.E. degree from Henan University, Kaifeng, China, in 2012 and the Ph.D. degree from the School of Computer Science and Technology, Tianjin University, Tianjin, China, in 2016. He is currently a postdoctoral in Department of Computer Science and Technology, Tsinghua University, Beijing, China. He got the China Scholarship Council Fellowship in 2014 and visited Washington University in St. Louis, USA, as a joint training student from Nov. 2014 to Nov. 2015. His current research interests include complex network analysis, machine learning, and data mining.