Deep learning is a popular method for monaural source separation, and especially for extracting a singing voice from a single-channel song. However, deep learning-based source separation ignores the geometrical struct...
详细信息
ISBN:
(纸本)9781509060672
Deep learning is a popular method for monaural source separation, and especially for extracting a singing voice from a single-channel song. However, deep learning-based source separation ignores the geometrical structure of the input data. This work develops a novel approach to source separation that is based on non-negativematrixfactorization (NMF) and deep recurrent neural networks (DRNN) with a locality-preserving constraint. First, NMF was used to learn patterns from training data. The learned patterns are linearly combined with the output of DRNN. Second, a locality preserving constraint is developed to exploit the inner structure of the input data in the DRNN learning process. Experimental results obtained using the MIR-1K dataset reveal that the proposed algorithm outperforms the baselines.
Fourier transform is the data processing naturally associated to most NMR experiments. Notable exceptions are Pulse Field Gradient and relaxation analysis, the structure of which is only partially suitable for FT. Wit...
详细信息
Fourier transform is the data processing naturally associated to most NMR experiments. Notable exceptions are Pulse Field Gradient and relaxation analysis, the structure of which is only partially suitable for FT. With the revamp of NMR of complex mixtures, fueled by analytical challenges such as metabolomics, alternative and more apt mathematical methods for data processing have been sought, with the aim of decomposing the NMR signal into simpler bits. Blind source separation is a very broad definition regrouping several classes of mathematical methods for complex signal decomposition that use no hypothesis on the form of the data. Developed outside NMR, these algorithms have been increasingly tested on spectra of mixtures. In this review, we shall provide an historical overview of the application of blind source separation methodologies to NMR, including methods specifically designed for the specificity of this spectroscopy. (C) 2014 Published by Elsevier B.V.
The purpose of this study was to find discriminating Raman spectral features between two major types of cancer, i.e., carcinoma and sarcoma. To this end, Raman spectra from adenocarcinoma, liposarcoma and fibrosarcoma...
详细信息
The purpose of this study was to find discriminating Raman spectral features between two major types of cancer, i.e., carcinoma and sarcoma. To this end, Raman spectra from adenocarcinoma, liposarcoma and fibrosarcoma samples were compared. A Raman system was used for the tissue Raman spectroscopic measurements at 785-nm laser excitation. After pre-processings, the Raman spectra were investigated, in major bands associated with protein and lipids, in the adenocarcinoma, liposarcoma, and fibrosarcoma groups. Principal component analysis and nonnegativematrixfactorization were performed for finding most significant features in discriminating the spectra of carcinoma from those of sarcoma samples. The findings of this study show that the lipid content in the sarcoma samples decreases compared with the carcinoma samples. The achieved accuracy in discriminating carcinoma from sarcoma by linear discriminant analysis is 93.75 % and 90.63 % using the first nine principal components and nonnegativematrixfactorization analysis, respectively.
non-negativematrixfactorization (NMF) has recently received a lot of attention in data mining, information retrieval, and computer vision. It factorizes a non-negative input matrix V into two non-negativematrix fac...
详细信息
ISBN:
(纸本)9781424452422
non-negativematrixfactorization (NMF) has recently received a lot of attention in data mining, information retrieval, and computer vision. It factorizes a non-negative input matrix V into two non-negativematrix factors V = WH such that W describes "clusters" of the datasets. Analyzing genotypes, social networks, or images, it can be beneficial to ensure V to contain meaningful "cluster centroids", i.e., to restrict W to be convex combinations of data points. But how can we run this convex NMF in the wild, i.e., given millions of data points? Triggered by the simple observation that each data point is a convex combination of vertices of the data convex hull, we propose to restrict W further to be vertices of the convex hull. The benefits of this convex-hull NMF approach are twofold. First, the expected size of the convex hull of, for example, n random Gaussian points in the plane is Omega(root log n), i.e., the candidate set typically grows much slower than the data set. Second, distance preserving low-dimensional embeddings allow one to compute candidate vertices efficiently. Our extensive experimental evaluation shows that convex-hull NMF compares favorably to convex NMF for large data sets both in terms of speed and reconstruction quality. Moreover, we show that our method can easily be applied to large-scale, real-world data sets, in our case consisting of 1.6 million images respectively 150 million votes on World of Warcraft (R) guilds.
This work addresses the task of tonic pitch identification in Indian classical music. The drone or the tambura establishes the tonic in Indian classical music. A cepstrum based pitch extraction technique is proposed t...
详细信息
ISBN:
(纸本)9781467359528;9781467359504
This work addresses the task of tonic pitch identification in Indian classical music. The drone or the tambura establishes the tonic in Indian classical music. A cepstrum based pitch extraction technique is proposed to identify the tuning of the tambura. We show that by identifying the musical note Sadja in the lower octave of a performance, the pitch of the tonic can be identified accurately. We also show that by estimating pitch of low energy frames, tonic can be identified with greater speed and higher accuracy. In order to further enhance the speed and also illustrate the ubiquitous nature of the tonic, a non-negativematrixfactorization (NMF) technique based method is developed to identify tonic. The proposed methods are validated by testing on a large varied dataset and accuracies close to 100% is reported.
Clustering is an unsupervised learning technique in that there is no explicit demarcation of data as training and test data. Clustering aims to group related records by measuring similarities among the attribute. Majo...
详细信息
ISBN:
(纸本)9781467327589;9781467327565
Clustering is an unsupervised learning technique in that there is no explicit demarcation of data as training and test data. Clustering aims to group related records by measuring similarities among the attribute. Major phase of clustering techniques is similarity measurement and it is based on different factors and parameters. The improved nonnegativematrixfactorization (NMF) based TCLUST (T-Clustering) algorithm is EM principle (Expectation Maximization) based algorithm, intended to search for approximate solutions. The EM algorithm is the efficient method of obtaining a solution to the mixture likelihood problem. Genes with a common function are often hypothesized to have correlated expression levels across different conditions. NMF clustering is introduced to find a small number of Meta genes, each defined as a positive linear combination of the genes in the expression data. The proposed clustering algorithm is applied to a genome scale gene expression dataset to enrichment analysis and to discover highly significant biological clusters.
This work addresses the concept of nonnegativematrixfactorization (NMF). Sonic relevant issues for its formulation as as a non-linear optimization problem will be discussed. The primary goal of NMI? is that of obtai...
详细信息
ISBN:
(纸本)9783642202667
This work addresses the concept of nonnegativematrixfactorization (NMF). Sonic relevant issues for its formulation as as a non-linear optimization problem will be discussed. The primary goal of NMI? is that of obtaining good quality approximations, namely for video/image visualization. The importance of the rank of the factor matrices and the use of global optimization techniques is investigated. Some computational experience is reported indicating that, in general, the relation between the quality of the obtained local minima and the factor matrices dimensions has a strong impact on the quality of the solutions associated with the decomposition.
暂无评论