Due to significant performance in the representation of data points, non-negative matrix factorization (NMF) has been widely applied in machine-learning fields, such as dimension reduction, image representation, featu...
详细信息
Due to significant performance in the representation of data points, non-negative matrix factorization (NMF) has been widely applied in machine-learning fields, such as dimension reduction, image representation, feature extraction, data mining, and so on. However, classical NMF suffers from a common issue, low efficiency in representing the internal geometric structure of data and sparsity limitation. To circumvent this problem, we innovatively propose a semi-supervised NMF algorithm called semi-supervised dual-graph regularization non-negative matrix factorization (LOSDNMF), into which dual-graph and bi-orthogonal constraints are embedded to reduce the inconsistency between the original matrix and the basic vectors while maintaining the manifold structures of the data and feature spaces. This strategy can fully explore the potential geometry information of the data, which is extremely beneficial to enhance the learning ability of the model. In addition, the local coordinate constraints are introduced to ensure good sparsity of the coefficient matrix and simplify the calculation. Furthermore, an iterative updating scheme for the optimization problem of LOSDNMF and its convergence proofs are also provided in detail. The effectiveness of the proposed method is verified on eight benchmark datasets. Experimental results show that our method can effectively improve clustering performance. (c) 2022 SPIE and IS&T
non-negative matrix factorization (NMF) has been introduced into the field of hyperspectral unmixing in the last ten years. To relieve the non-convex problem of NMF, different constraints are imposed on NMF. In this p...
详细信息
non-negative matrix factorization (NMF) has been introduced into the field of hyperspectral unmixing in the last ten years. To relieve the non-convex problem of NMF, different constraints are imposed on NMF. In this paper, a new constraint, termed the endmember dissimilarity constraint (EDC), is proposed. The proposed constraint can measure the difference between the signatures as well as constrain the signatures to be smooth. A set of smooth spectra contained in the dataset space with the largest differences can be obtained, as far as is possible, which can be seen as endmembers. The experimental performances of our method and other state-of-the-art constrained NMF algorithms were obtained and analyzed, proving that the proposed method outperforms other NMF unmixing methods.
We compared non-negative matrix factorization (NMF) and convolution kernel compensation techniques for high-density electromyogram decomposition. The experimental data were recorded from nine healthy persons during co...
详细信息
We compared non-negative matrix factorization (NMF) and convolution kernel compensation techniques for high-density electromyogram decomposition. The experimental data were recorded from nine healthy persons during controlled single degree of freedom (DOF) wrist flexion-extension, supination-pronation, and ulnar-radial deviation movements. We assembled the identified motor units and NMF components into three groups. Those active mostly during the first and the second movement direction per DOF were placed in the G1 and G3 groups, respectively. The remaining components were nonspecific for movement direction and were placed in the G2 group. In ulnar and radial deviation, the relative energies of identified cumulative motor unit spike trains (CSTs) and NMF components were similarly distributed among the groups. In other two movement types, the energy of NMF components in the G2 group was significantly larger than the energy of CSTs. We further performed a coherence analysis between CSTs and sums of NMF components in each group. Both decompositions demonstrated a solid match, but only at frequencies <3 Hz. At higher frequencies, the coherence hardly exceeded the value of 0.5. Potential reasons for these discrepancies include the negative impact of motor unit action potential shapes and noise on NMF decomposition.
Network is an abstract expression of subjects and the relationships among them in the real-world system. Research on community detection can help people understand complex systems and identify network functionality. I...
详细信息
Network is an abstract expression of subjects and the relationships among them in the real-world system. Research on community detection can help people understand complex systems and identify network functionality. In this paper, we present a novel approach to community detection that utilizes a nonnegativematrixfactorization (NMF) model to divide overlapping community from networks. The study is based on the different physical meanings of the pair of matrices W and H to optimize the constraint condition. Many community detection algorithms based on NMF require the number of known communities as a prior condition, which limits the field of application of the algorithms. This paper handled the problem by feature matrix preprocessing and ranking optimization, so that the proposed algorithm can divide the network structure with unknown community number. Experiments demonstrated that the proposed algorithm can effectively divide the community structure, and identify network overlay communities and overlapping nodes.
Time-frequency (TF) representation has found wide use in many challenging signal processing tasks including classification, interference rejection, and retrieval. Advances in TF analysis methods have led to the develo...
详细信息
Time-frequency (TF) representation has found wide use in many challenging signal processing tasks including classification, interference rejection, and retrieval. Advances in TF analysis methods have led to the development of powerful techniques, which use non-negative matrix factorization (NMF) to adaptively decompose the TF data into TF basis components and coefficients. In this paper, standard NMF is modified for TF data, such that the improved TF bases can be used for signal classification applications with overlapping classes and data retrieval. The new method, called jointly learnt NMF (JLNMF) method, identifies both distinct and shared TF bases and is able to use the decomposed bases to successfully retrieve and separate the class-specific information from data. The paper provides the framework of the proposed JLNMF cost function and proposes a projected gradient framework to solve for limit point stationarity solutions. The developed algorithm has been applied to a synthetic data retrieval experiment and epileptic spikes in EEG signals of infantile spasms and discrimination of pathological voice disorder. The experimental results verified that JLNMF successfully identified the class-specific information, thus enhancing data separation performance.
In last decade space-density of monitoring stations increased, in to addition also air pollution modeling made big progress. Using diversity of big data can lead to better knowledge about air pollution at continental ...
详细信息
In last decade space-density of monitoring stations increased, in to addition also air pollution modeling made big progress. Using diversity of big data can lead to better knowledge about air pollution at continental scale. The focus of presented study is the data-driven approach using non-negative matrix factorization to provide new insights and to study the characteristic space-time particulate-matter patterns across Europe. We analyzed the PM10 concentrations obtained from 1097 monitoring stations (AirBase data) and the Monitoring Atmospheric Composition and Climate (MACC) modeled fields for a period of 3 years. We distinguished five characteristic patterns obtained from the AirBase data and five patterns from the MACC data. A comparison between the AirBase and MACC data shows a good spatial overlap for the east Europe, central Europe and the Mediterranean patterns. However, it should be noted that an analysis of the MACC data revealed two additional marine patterns: the Celtic and the North Seas. The Po Valley and Balkan patterns were very clearly identified when analyzing the AirBase data. In order to better understand the influence of the synoptic situation on the particulate-matter concentrations the synoptic meteorological situations were additionally analyzed. The cold season, low wind and very stable conditions, which can last for several days, is the most common situation linked to high concentrations of anthropogenic air pollution with particulate matter. In contrast, for the Mediterranean pattern the most common situation (high factor loadings) is observed during the summer period. This pattern also exhibits a clearer annual cycle. A closer look at the sea-salt patterns (Celtic and North Seas) shows low time-series correlations between these two factors. Nevertheless, the physical mechanism is the same: a steep gradient between the cyclone and the anti-cyclone that causes high winds and, consequently, higher sea-salt production. (C) 2016 Elsevier Ltd. A
A multimodal voice conversion (VC) method for noisy environments is proposed. In our previous non-negative matrix factorization (NMF)-based VC method, source and target exemplars are extracted from parallel training d...
详细信息
A multimodal voice conversion (VC) method for noisy environments is proposed. In our previous non-negative matrix factorization (NMF)-based VC method, source and target exemplars are extracted from parallel training data, in which the same texts are uttered by the source and target speakers. The input source signal is then decomposed into source exemplars, noise exemplars, and their weights. Then, the converted speech is constructed from the target exemplars and the weights related to the source exemplars. In this study, we propose multimodal VC that improves the noise robustness of our NMF-based VC method. Furthermore, we introduce the combination weight between audio and visual features and formulate a new cost function to estimate audio-visual exemplars. Using the joint audio-visual features as source features, VC performance is improved compared with that of a previous audio-input exemplar-based VC method. The effectiveness of the proposed method is confirmed by comparing its effectiveness with that of a conventional audio-input NMF-based method and a Gaussian mixture model-based method.
An unsupervised single channel audio separation method from pattern recognition viewpoint is presented. The proposed method does not require training knowledge and the separation system is based on non-uniform time-fr...
详细信息
An unsupervised single channel audio separation method from pattern recognition viewpoint is presented. The proposed method does not require training knowledge and the separation system is based on non-uniform time-frequency (TF) analysis and feature extraction. Unlike conventional research that concentrates on the use of spectrogram or its variants, the proposed separation algorithm uses an alternative TF representation based on the gammatone filterbank. In particular, the monaural mixed audio signal is shown to be considerably more separable in this non-uniform TF domain. The analysis of signal separability to verify this finding is provided. In addition, a variational Bayesian approach is derived to learn the sparsity parameters for optimizing the matrixfactorization. Experimental tests have been conducted, which show that the extraction of the spectral dictionary and temporal codes is more efficient using sparsity learning and subsequently leads to better separation performance. (C) 2014 Acoustical Society of America.
This paper describes a novel two-stage dereverberation feature enhancement method for noise-robust automatic speech recognition. In the first stage, an estimate of the dereverberated speech is generated by matching th...
详细信息
This paper describes a novel two-stage dereverberation feature enhancement method for noise-robust automatic speech recognition. In the first stage, an estimate of the dereverberated speech is generated by matching the distribution of the observed reverberant speech to that of clean speech, in a decorrelated transformation domain that has a long temporal context in order to address the effects of reverberation. The second stage uses this dereverberated signal as an initial estimate within a non-negative matrix factorization framework, which jointly estimates a sparse representation of the clean speech signal and an estimate of the convolutional distortion. The proposed feature enhancement method, when used in conjunction with automatic speech recognizer back-end processing, is shown to improve the recognition performance compared to three other state-of-the-art techniques.
non-negative matrix factorization (NMF) as a popular technique for finding parts-based, linear representations of non-negative data has been successfully applied in a wide range of applications, such as feature learni...
详细信息
non-negative matrix factorization (NMF) as a popular technique for finding parts-based, linear representations of non-negative data has been successfully applied in a wide range of applications, such as feature learning, dictionary learning, and dimensionality reduction. However, both the local manifold regularization of data and the discriminative information of the available label have not been taken into account together in NMF. We propose a new semisupervised matrix decomposition method, called manifold regularized non-negative matrix factorization (MRNMF) with label information, which incorporates the manifold regularization and the label information into the NMF to improve the performance of NMF in clustering tasks. We encode the local geometrical structure of the data space by constructing a nearest neighbor graph and enhance the discriminative ability of different classes by effectively using the label information. Experimental comparisons with the state-of-the-art methods on theCOIL20, PIE, Extended Yale B, and MNIST databases demonstrate the effectiveness of MRNMF. (C) 2016 SPIE and IS&T
暂无评论