Hyperspectral imagery generally contains a very large amount of data due to hundreds of spectral *** selection is often applied firstly to reduce computational cost and facilitate subsequent tasks such as land-cover c...
详细信息
Hyperspectral imagery generally contains a very large amount of data due to hundreds of spectral *** selection is often applied firstly to reduce computational cost and facilitate subsequent tasks such as land-cover classification and higher level image *** this paper,we propose a new band selection algorithm using sparse nonnegative matrix factorization (sparse NMF).Though acting as a clustering method for band selection,sparse NMF need not consider the distance metric between different spectral bands,which is often the key step for most common clustering-based band selection *** imposing sparsity on the coefficient matrix,the bands' clustering assignments can be easily indicated through the largest entry in each column of the *** results showed that sparse NMF provides considerable insight into the clustering-based band selection problem and the selected bands are good for land-cover classification.
At present, the mining and analysis of teaching data is mainly aimed at the online courses data, but not mixed data, which is fused by the traditional offline-classroom and online teaching data. Meanwhile, the most ev...
详细信息
ISBN:
(纸本)9781450388894
At present, the mining and analysis of teaching data is mainly aimed at the online courses data, but not mixed data, which is fused by the traditional offline-classroom and online teaching data. Meanwhile, the most evaluation models are constructed by the learning data to evaluate the teaching quality of teachers, but not to evaluate and grade the individual quality of students. In fact, the evaluation and grading of students' quality can effectively provide more targeted teaching intervention for students of different levels based on the data analysis. To address these issues, the online teaching data is fused by the students' learning behavior data of traditional course to form the mixed data in this paper, and then the sparse non-negative matrixfactorization (SNMF) method is adopted to extract the feature clusters of mixed learning data. According to the weights of the extracted cluster features, the multi-level feature indicators are selected in turn to construct the hierarchical evaluation index system. Finally, the comprehensive weighting method is adopted to evaluate and grade the individual students. In this paper, the mixed teaching data of computer basic course of our school is formed, and then the weights of feature clusters are calculated by SNMF and an evaluation model is established to evaluate and grade the students. The grading results are in accordance with the normal distribution and basically consistent with the grading distribution of students' final examination scores. Thus the validity of the model and method proposed in this paper is proved.
In this paper we study nonconvex and nonsmooth optimization problems with semialgebraic data, where the variables vector is split into several blocks of variables. The problem consists of one smooth function of the en...
详细信息
In this paper we study nonconvex and nonsmooth optimization problems with semialgebraic data, where the variables vector is split into several blocks of variables. The problem consists of one smooth function of the entire variables vector and the sum of nonsmooth functions for each block separately. We analyze an inertial version of the proximal alternating linearized minimization algorithm and prove its global convergence to a critical point of the objective function at hand. We illustrate our theoretical findings by presenting numerical experiments on blind image deconvolution, on sparse nonnegative matrix factorization and on dictionary learning, which demonstrate the viability and effectiveness of the proposed method.
We introduce a proximal alternating linearized minimization (PALM) algorithm for solving a broad class of nonconvex and nonsmooth minimization problems. Building on the powerful Kurdyka-Aojasiewicz property, we derive...
详细信息
We introduce a proximal alternating linearized minimization (PALM) algorithm for solving a broad class of nonconvex and nonsmooth minimization problems. Building on the powerful Kurdyka-Aojasiewicz property, we derive a self-contained convergence analysis framework and establish that each bounded sequence generated by PALM globally converges to a critical point. Our approach allows to analyze various classes of nonconvex-nonsmooth problems and related nonconvex proximal forward-backward algorithms with semi-algebraic problem's data, the later property being shared by many functions arising in a wide variety of fundamental applications. A by-product of our framework also shows that our results are new even in the convex setting. As an illustration of the results, we derive a new and simple globally convergent algorithm for solving the sparse nonnegative matrix factorization problem.
We propose the employment of nonnegativesparse linear feature extraction as a tool for unsupervised spectral unmixing sparse feature extraction can be seen as a general linear unmixing approach that maps the data int...
详细信息
ISBN:
(纸本)9781424433940
We propose the employment of nonnegativesparse linear feature extraction as a tool for unsupervised spectral unmixing sparse feature extraction can be seen as a general linear unmixing approach that maps the data into a new dimensional space in which each of the components has only a limited number of non-zero values Unlike other transforms that target decorrelation or statistical independence, our focus is on the enforcement of sparseness by imposing restrictions (such as cardinality or norm relationships), as well as nonnegativity. When compared to the linear mixing model, the sparse components can be naturally associated to the abundance of endmembers, and the inverse transform to the endmembers. Our approach is a variant of a well known technique based on nonnegativematrixfactorization (NMF). In most of the cases. the NMF components are produced using a gradient descent optimization algorithm that was previously shown to converge. To validate our approach we use quantitative (classification) and qualitative (visualization) analysis of hyperspectral data sets.
nonnegativematrixfactorization (NMF) has proven to be a useful tool for the analysis of nonnegative multivariate data. However, it is known not to lead to unique results when applied to blind source separation (BSS)...
详细信息
nonnegativematrixfactorization (NMF) has proven to be a useful tool for the analysis of nonnegative multivariate data. However, it is known not to lead to unique results when applied to blind source separation (BSS) problems. In this paper we present an extension of NMF capable of solving the BSS problem when the underlying sources are sufficiently sparse. In contrast to most well-established BSS methods, the devised algorithm is capable of solving the BSS problem in cases where the underlying sources are not independent or uncorrelated. As the proposed fitness function is discontinuous and possesses many local minima, we use a genetic algorithm for its minimization. Finally, we apply the devised algorithm to real world microarray data. (C) 2008 Elsevier B.V. All rights reserved.
In this study we focus on diagnostic classification tasks and the extraction of related marker genes from gene expression profiles. We apply ICA and sparse NMF to various microarray data sets. The latter monitor the g...
详细信息
ISBN:
(纸本)9781424408290
In this study we focus on diagnostic classification tasks and the extraction of related marker genes from gene expression profiles. We apply ICA and sparse NMF to various microarray data sets. The latter monitor the gene expression levels of either human breast cancer (HBC) cell lines [1] or the famous leucemia data set [2] under various environmental conditions. We show that these matrix decomposition techniques are able to identify relevant signatures in the deduced matrices and extract marker genes from these gene expression profiles. With these marker genes corresponding test data sets can be classified into related diagnostic categories.
The decomposition of surface electromyogram data sets (s-EMG) is studied using blind source separation techniques based on sparseness;namely independent component analysis, sparse nonnegative matrix factorization, and...
详细信息
The decomposition of surface electromyogram data sets (s-EMG) is studied using blind source separation techniques based on sparseness;namely independent component analysis, sparse nonnegative matrix factorization, and sparse component analysis. When applied to artificial signals we find noticeable differences of algorithm performance depending on the source assumptions. In particular, sparse nonnegative matrix factorization outperforms the other methods with regard to increasing additive noise. However, in the case of real s-EMG signals we show that despite the fundamental differences in the various models, the methods yield rather similar results and can successfully separate the source signal. This can be explained by the fact that the different sparseness assumptions (super-Gaussianity, positivity together with minimal l-norm and fixed number of zeros, respectively) are all only approximately fulfilled thus apparently forcing the algorithms to reach similar results, but from different initial conditions. (C) 2005 Elsevier B.V. All rights reserved.
暂无评论