Brief PapersNonparametric discriminant multi-manifold learning for dimensionality reduction
Introduction
During last decade, dimensionality reduction has been attracting considerable attention in many fields as pattern recognition, data mining, computer vision and machine learning. As a fundamental problem in these fields, dimensionality reduction plays an important role in data analysis with the goal to find a meaningful low dimensional representation of high dimensional data. In regard to pattern recognition, dimensionality reduction is an effective and feasible way to overcome the “curse of dimensionality”. Moreover, since there are large volumes of high dimensional data in numerous real world applications, some of which are perhaps superfluous, so extracting the most useful features from the real world data using dimensionality reduction techniques not only helps to explore the essential structure of the original data, but also contributes to accomplish the task of classification at low computational cost.
Dimensionality reduction methods can be categorized into linear models and nonlinear ones. Many linear methods such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Independent Component Analysis (ICA) have been widely used on practical applications with good performance [34]. However, they also expose their limitations when applied to nonlinear distributed data. So in the last few years, some approaches including manifold learning have been developed for nonlinear dimensionality reduction. Among all the manifold learning methods, Isometric Mapping (ISOMAP) [1], Locally Linear Embedding (LLE) [2], [3], Laplacian Eigenmaps (LE) [4], Local Tangent Space Alignment (LTSA) [5], Maximum Variance Unfolding (MVU) [6] and Riemannian Manifold Learning (RML) [7] are their representatives. It was shown by many examples that these methods have yielded impressive results on artificial and real world data sets. Compared to other nonlinear dimensionality reduction methods, manifold learning shows its superiorities in the following. On the one hand, manifold learning can explore the essential dimensions of manifold data embedded in high dimensional space. For instance, in Ref. [1], [2], the left–right pose, up–down pose and lighting direction can be found to three essential dimensions of face manifold. On the other hand, manifold learning pursuits to embed the original data with high dimensions in a lower dimensional space by locality preserving, where the locality can be approached using k Nearest Neighbors (KNN) criterion. Thus manifold learning can be efficiently applied for data visualization. The reason is concluded that both manifold learning and data visualization have the similar goal to project the original data into a 2-D or 3-D space, where their intrinsic structure information will be preserved as much as possible.
However, when employed to data classification, manifold learning appears some shortcomings. Firstly, manifold learning yields embeddings directly based on training data sets. Owing to the implicitness of nonlinear mapping, for any new test sample, the original manifold learning methods cannot easily obtain its projection in the embedding space by utilizing low-dimensional embeddings of training sets, which greatly confines applications of the original manifold learning algorithms to pattern classification. In order to overcome this out-of-sample problem [8], linearization, kernelization, tensorization and other tricks were proposed, which were validated efficient to find the low dimensional embeddings of test data on the basis of the mapping results of training samples [9], [10].
Secondly, it is unproblematic for the existing manifold learning algorithms as they try to approach a simple manifold. However, if there are many manifolds, how to clearly identify different manifolds still needs further demonstrations. For example, if face images sampled from many persons do exist in a high dimensional space, then different persons׳ face images should lie on the corresponding manifolds. So it is necessary to distinguish individual face images from different manifolds. In order to achieve optimal recognition results, the recovered embeddings associated to different manifolds should be as separable as possible in the final embedding space, which poses a problem that might be called “classification-oriented multi-manifold learning” [11]. The problem cannot be solved by some current manifold learning algorithms and their supervised versions [12], [13], [14], [15], [16], [17], [18], [19], [20] because they all just concentrate on the characterization of “locality” and do not take into account the variances among manifolds. Recently, researchers put forward some discriminant multi-manifold learning methods, where supervised graph was constructed to approach manifolds distances. Lai proposed a soft margin scatter and a soft within-class scatter, both of which were introduced to find an optimal subspace for data classification [21]. Wang presented a maximum inter-class and marginal discriminant embedding for manifolds identification [22]. On the basis of Unsupervised Discriminant Projection (UDP) [11], a locally statistical uncorrelation was advanced by Chen as a constraint [23]. Later, Lu defined an inter-manifold graph and an intra-manifold graph according to class information and then the corresponding graph Laplacian spectrum was used to search the optimal projection [24]. Similar to Lu, Chen adopted the least reconstruction trick to both inter-class graph and intra-class graph to seek a discirminant subspace [25]. All the methods mentioned above characterized the separability of manifolds globally and failed to take advantage of the local distances between manifolds.
So in this paper, a dimensionality reduction method, named Nonparametric Discirminant Multi-manifold Learning (NDML), is presented and involved in multiply manifolds identification. In the proposed NDML algorithm, a novel nonparametric manifold-to-manifold distance is defined to model separability between manifolds, where both labels and local structure information of manifolds are all considered. Moreover, the linearization trick is also introduced to avoid out-of-sample problem. At last, an objective function is constructed to explore an optimal subspace with the maximum manifold-to-manifold distances and the minimum manifolds locality.
The rest of the paper is organized as follows: Section 2 simply reviews LLE and Constrained Maximum Variance Mapping (CMVM) [26]. In Section 3, the principle of NDML is addressed in details. Experimental results on AR face data, ORL face data and YaleB face data are offered in Section 4 and the paper is finished with some conclusions in Section 5.
Section snippets
Reviews of LLE and CMVM
There are many dimensionality reduction methods related to the proposed NDML such as LLE and CMVM, which will be briefly reviewed in the following.
Nonparametric discirminant multi-manifold learning
Accompanying with the classical manifold learning methods, more and more supervised extensions are booming for dimensionality reduction. Some take advantage of class information to adjust neighborhood weights in KNN graph; others are integrated to LDA [27] for discriminant dimensionality reduction. However, most of these supervised versions pay more attention to constructions of local graph instead of local distance between manifolds, which can be introduced to measure the separability of
Experiments
In this section, experiments will be conducted on some benchmark data sets including AR face data, ORL face data and YaleB face data, where UDP, CMVM, LDA and the proposed NDML algorithm are all employed to reduce dimensionality of the original face data. Moreover, in the low dimensional subspace, the Nearest Neighbor (NN) classifier is also adopted, by which the labels of those test data will be predicted.
However, when carrying out experiments, KNN is applied to approach the locality in UDP,
Conclusion
In this paper, a nonparametric discriminant multi-manifold learning method is proposed for dimensionality reduction. In the proposed method, manifolds distance is locally or nonparametric defined which can be modeled to distance between any point and mean of its inter-class k nearest neighbors. Moreover, an objective function is constructed to explore the low dimensional subspace with the maximum manifolds distance and the minimum locality, which is validated either by the theoretical analysis
Acknowledgment
This work was partly supported by the Grants of the National Natural Science Foundation of China (61273303, 61273225, 61373109 and 61472280), the China Postdoctoral Science Foundation (20100470613 and 201104173), the Natural Science Foundation of Hubei Province (2010CDB03302), the Research Foundation of Education Bureau of Hubei Province (Q20121115), the Program of Wuhan Subject Chief Scientist (201150530152), the Hong Kong Scholars Program (XJ2012012) and the Open Project Program of the
Bo Li received his M.Sc. and Ph.D. degree in Mechanical and Electronic Engineering from Wuhan University of Technology in 2003, Pattern Recognition and Intelligent System from University of Science and Technology of China in 2008, respectively. Now, he is an associated professor at School of Computer Science and Technology, Wuhan University of Science and Technology. He is also a research associated in Ryerson University. His research interests include machine learning, pattern recognition,
References (34)
- et al.
Supervised locally linear embedding with probability-based distance for classification
Comput. Math. Appl.
(2009) - et al.
Locally linear discriminant embedding: an efficient method for face recognition
Pattern Recognit.
(2008) - et al.
Discriminant subspace learning constrained by locally statistical uncorrelation for face recognition
Neural Netw.
(2013) - et al.
Recontructive discriminant analysis: a feature extraction method induced from linear regression classification
Neurocomputing
(2012) - et al.
A global geometric framework for nonlinear dimensionality reduction
Science
(2000) - et al.
Nonlinear dimensionality reduction by locally linear embedding
Science
(2000) - et al.
Think globally, fit locally: unsupervised learning of low dimensional manifolds
J. Mach. Learn. Res.
(2003) - et al.
Laplacian eigenmaps for dimensionality reduction and data representation
Neural Comput.
(2003) - et al.
Principal manifolds and nonlinear dimensionality reduction by local tangent space alignment
SIAM J. Sci. Comput.
(2004) - et al.
Unsupervised learning of image manifolds by semi-definite programming
Int. J. Comput. Vis.
(2006)
Riemannian manifold learning
IEEE Trans. Pattern Anal. Mach. Intell.
Out-of-sample extensions for LLE, Isomap, MDS, eigenmaps, and spectral clustering
Technical Report 1238
Graph embedding: a general framework for dimensionality reduction
IEEE Trans. Pattern Anal. Mach. Intell.
Globally maximizing, locally minimizing: unsupervised discriminant projection with application to face and palm biometrics
IEEE Trans. Pattern Anal. Mach. Intell.
Cited by (30)
Polygonal Coordinate System: Visualizing high-dimensional data using geometric DR, and a deterministic version of t-SNE
2021, Expert Systems with ApplicationsCitation Excerpt :KPCA is a nonlinear method that transforms a nonlinear problem into a linear problem in some embedded feature space. Nonlinear DR algorithms include, in addition to KPCA, various manifold learning techniques (Li, Li, & Zhang, 2015; Paul & Chalup, 2017; Holiday et al., 2019). Without the need for external information (e.g., class labels), manifold learning is a nonlinear approach to DR that considers the generalization of nonlinear data structures.
A survey of CAPTCHA technologies to distinguish between human and computer
2020, NeurocomputingA pseudo-dynamic search ant colony optimization algorithm with improved negative feedback mechanism
2020, Cognitive Systems ResearchA survey on Laplacian eigenmaps based manifold learning methods
2019, NeurocomputingSemi-supervised metric learning in stratified spaces via intergrating local constraints and information-theoretic non-local constraints
2018, NeurocomputingCitation Excerpt :Both groups of metric learning methods enforce some false prior knowledge which is not satisfied in the intersecting points of manifolds, and therefore, the existing metric learning methods are not suitable for the stratified space. Many studies have shown that data could be better modeled by the assumption that lies on the stratified space [1,16,19,23,28,52]. The stratified space is a mixture of manifolds representing different characteristics and complexities of the data set.
Robust locally linear embedding algorithm for machinery fault diagnosis
2018, NeurocomputingCitation Excerpt :Recently, many researchers have devoted much time to extracting the important feature of a data set by manifold learning algorithms. Manifold learning algorithms [5–8] can project an original data set into a low-dimensional feature space, where the types of each data can be easily recognized [9]. In general, manifold learning algorithms can be roughly divided into two categories: linear methods [10,11] and non-linear methods [12,13].
Bo Li received his M.Sc. and Ph.D. degree in Mechanical and Electronic Engineering from Wuhan University of Technology in 2003, Pattern Recognition and Intelligent System from University of Science and Technology of China in 2008, respectively. Now, he is an associated professor at School of Computer Science and Technology, Wuhan University of Science and Technology. He is also a research associated in Ryerson University. His research interests include machine learning, pattern recognition, image processing and bioinformatics.
Jun Li received his M.Sc. in Computer Science from Wuhan University of Technology in 2003. Now, he is an associated professor at School of Computer Science and Technology, Wuhan University of Science and Technology. Moreover, he is also a doctoral candidate in School of Computer, Wuhan University. His research interests include machine learning, pattern recognition, image processing and intelligent computing.
Xiao-Ping Zhang (M׳97, SM׳02) received B.S. and Ph.D. degrees from Tsinghua University, in 1992 and 1996, respectively, both in Electronic Engineering. He holds an MBA in Finance, Economics and Entrepreneurship with Honors from the University of Chicago Booth School Of Business, Chicago, IL. He is a professor in Ryerson University. He is currently an Associate Editor for IEEE Transactions on Signal Processing, IEEE Transactions on Multimedia, IEEE Signal Processing letters and for Journal of Multimedia. His research interests include intelligent computing, bioinformatics, multimedia communication and signal processing.