Search

In this paper, we propose a new adaptive cross algorithm for computing a low tubal rank approximation of third-order tensors, with less memory usage and lower computational complexity than the truncated tensor SVD (t-SVD). This makes it applicable for decomposing large-scale tensors. We conduct numerical experiments on synthetic and real-world datasets to confirm the efficiency and feasibility of the proposed algorithm. The simulation results show more than one order of magnitude acceleration in the computation of low tubal rank approximations for large-scale tensors. An application to pedestrian attribute recognition is also presented. [ABSTRACT FROM AUTHOR]

This paper presents the neural dynamical network to compute a best rank-one approximation of a real-valued tensor. We implement the neural network model by the ordinary differential equations (ODE), which is a class of continuous-time recurrent neural network. Several new properties of solutions for the neural network are established. We prove that the locally asymptotic stability of solutions for ODE by constructive an appropriate Lyapunov function under mild conditions. Furthermore, we also discuss how to use the proposed neural networks for solving the tensor eigenvalue problem including the tensor H-eigenvalue problem, the tensor Z-eigenvalue problem, and the generalized eigenvalue problem with symmetric-definite tensor pairs. Finally, we generalize the proposed neural networks to the computation of the restricted singular values and the associated restricted singular vectors of real-valued tensors. We illustrate and validate theoretical results via numerical simulations. [ABSTRACT FROM AUTHOR]

Highlights: [•] Noninvasive diagnosis of melanoma from clinical color image. [•] Novel feature extraction based on tensor decomposition and nonlinear transformation. [•] Interesting alternative and competitive performance in comparison with state-of-the-art methods. [•] Diagnosis from color RGB image possibly enabling design of cost-effective prescreening devices. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we provide two generalizations of the CUR matrix decomposition (also known as pseudo-skeleton approximation method ) to the case of -way arrays (tensors). These generalizations, which we called Fiber Sampling Tensor Decomposition types 1 and 2 (FSTD1 and FSTD2), provide explicit formulas for the parameters of a rank- Tucker representation (the core tensor of size and the matrix factors of sizes , ) based only on some selected entries of the original tensor. FSTD1 uses -mode fibers of the original tensor while FSTD2 uses exactly fibers in each mode as matrix factors, as suggested by the existence theorem provided in Oseledets et al. (2008) , with a core tensor defined in terms of the entries of a subtensor of size . For our results are reduced to the already known CUR matrix decomposition where the core matrix is defined as the inverse of the intersection submatrix, i.e. . Additionally, we provide an adaptive type algorithm for the selection of proper fibers in the FSTD1 model which is useful for large scale applications. Several numerical results are presented showing the performance of our FSTD1 Adaptive Algorithm compared to two recently proposed approximation methods for 3-way tensors. [Copyright &y& Elsevier]

Abstract: Objective: Development of an EEG preprocessing technique for improvement of detection of Alzheimer''s disease (AD). The technique is based on filtering of EEG data using blind source separation (BSS) and projection of components which are possibly sensitive to cortical neuronal impairment found in early stages of AD. Methods: Artifact-free 20s intervals of raw resting EEG recordings from 22 patients with Mild Cognitive Impairment (MCI) who later proceeded to AD and 38 age-matched normal controls were decomposed into spatio-temporally decorrelated components using BSS algorithm ‘AMUSE’. Filtered EEG was obtained by back projection of components with the highest linear predictability. Relative power of filtered data in delta, theta, alpha 1, alpha 2, beta 1, and beta 2 bands were processed with Linear Discriminant Analysis (LDA). Results: Preprocessing improved the percentage of correctly classified patients and controls computed with jack-knifing cross-validation from 59 to 73% and from 76 to 84%, correspondingly. Conclusions: The proposed approach can significantly improve the sensitivity and specificity of EEG based diagnosis. Significance: Filtering based on BSS can improve the performance of the existing EEG approaches to early diagnosis of Alzheimer''s disease. It may also have potential for improvement of EEG classification in other clinical areas or fundamental research. The developed method is quite general and flexible, allowing for various extensions and improvements. [Copyright &y& Elsevier]

Most of source separation methods focus on stationary sources, so higher-order statistics is necessary for successful separation, unless sources are temporally correlated. For nonstationary sources, however, it was shown [Neural Networks 8 (1995) 411] that source separation could be achieved by second-order decorrelation. In this paper, we consider the cost function proposed by Matsuoka et al. [Neural Networks 8 (1995) 411] and derive natural gradient learning algorithms for both fully connected recurrent network and feedforward network. Since our algorithms employ the natural gradient method, they possess the equivariant property and find a steepest descent direction unlike the algorithm [Neural Networks 8 (1995) 411]. We also show that our algorithms are always locally stable, regardless of probability distributions of nonstationary sources. [Copyright &y& Elsevier]

In this paper, we formulate a mixed-integer problem for sparse signal reconstruction and reformulate it as a global optimization problem with a surrogate objective function subject to underdetermined linear equations. We propose a sparse signal reconstruction method based on collaborative neurodynamic optimization with multiple recurrent neural networks for scattered searches and a particle swarm optimization rule for repeated repositioning. We elaborate on experimental results to demonstrate the outperformance of the proposed approach against ten state-of-the-art algorithms for sparse signal reconstruction. [ABSTRACT FROM AUTHOR]

Abstract: This paper investigates whether some well understood principles of human behavioral analysis can be used to design novel paradigms for affective brain–computer/machine interfaces. This is achieved by using the visual, audio, and audiovisual stimuli representing human emotions. The analysis of brain responses to such stimuli involves several challenges related to the conditioning of brain electrical responses, extraction of the responses to stimuli and mutual information between the several physiological recording modalities used. This is achieved in the time–frequency domain, using multichannel empirical mode decomposition (EMD), which proves very accurate in the joint analysis of neurophysiological and peripheral body signals. Our results indicate the usefulness of such an approach and confirm the possibility of using affective brain–computer/machine interfaces. [Copyright &y& Elsevier]

Abstract: This study pursues the optimization of the brain responses to small reversing patterns in a Steady-State Visual Evoked Potentials (SSVEP) paradigm, which could be used to maximize the efficiency of applications such as Brain–Computer Interfaces (BCI). We investigated the SSVEP frequency response for 32 frequencies (5–84Hz), and the time dynamics of the brain response at 8, 14 and 28Hz, to aid the definition of the optimal neurophysiological parameters and to outline the onset-delay and other limitations of SSVEP stimuli in applications such as our previously described four-command BCI system. Our results showed that the 5.6–15.3Hz pattern reversal stimulation evoked the strongest responses, peaking at 12Hz, and exhibiting weaker local maxima at 28 and 42Hz. After stimulation onset, the long-term SSVEP response was highly non-stationary and the dynamics, including the first peak, was frequency-dependent. The evaluation of the performance of a frequency-optimized eight-command BCI system with dynamic neurofeedback showed a mean success rate of 98%, and a time delay of 3.4s. Robust BCI performance was achieved by all subjects even when using numerous small patterns clustered very close to each other and moving rapidly in 2D space. These results emphasize the need for SSVEP applications to optimize not only the analysis algorithms but also the stimuli in order to maximize the brain responses they rely on. [Copyright &y& Elsevier]

Abstract: We propose a novel similarity measure, called the correntropy coefficient, sensitive to higher order moments of the signal statistics based on a similarity function called the cross-correntopy. Cross-correntropy nonlinearly maps the original time series into a high-dimensional reproducing kernel Hilbert space (RKHS). The correntropy coefficient computes the cosine of the angle between the transformed vectors. Preliminary experiments with simulated data and multichannel electroencephalogram (EEG) signals during behaviour studies elucidate the performance of the new measure versus the well-established correlation coefficient. [Copyright &y& Elsevier]

In this paper we propose a new approach to blind separation of independent source signals that, while avoiding the imposition of an orthogonal mixing matrix, is robust with respect to the existence of additive Gaussian noise in the mixture. We demonstrate that, for the wide class of source distributions with certain non-null cumulants and a pre-specified scaling, separation is always a saddle point of a cumulant-based cost function. We propose a quasi-Newton approach for determining this saddle point. This enables us to obtain a family of separation algorithms which, based on higher order statistics, yields unbiased estimates even in the presence of large Gaussian noise and has the interesting property of local isotropic convergence. Another family of algorithms that incorporates second-order statistics loses the former desirable convergence properties but it provides more precise estimates in the absence of noise. Extensive computer simulations confirm robustness and the excellent performance of the resulting algorithms. [Copyright &y& Elsevier]

Multi-channel EEG data are usually necessary for spatial pattern identification in motor imagery (MI)-based brain computer interfaces (BCIs). To some extent, signals from some channels containing redundant information and noise may degrade BCI performance. We assume that the channels related to MI should contain common information when participants are executing the MI tasks. Based on this hypothesis, a correlation-based channel selection (CCS) method is proposed to select the channels that contained more correlated information in this study. The aim is to improve the classification performance of MI-based BCIs. Furthermore, a novel regularized common spatial pattern (RCSP) method is used to extract effective features. Finally, a support vector machine (SVM) classifier with the Radial Basis Function (RBF) kernel is trained to accurately identify the MI tasks. An experimental study is implemented on three public EEG datasets (BCI competition IV dataset 1, BCI competition III dataset IVa and BCI competition III dataset IIIa) to validate the effectiveness of the proposed methods. The results show that the CCS algorithm obtained superior classification accuracy (78% versus 56.4% for dataset1, 86.6% versus 76.5% for dataset 2 and 91.3% versus 85.1% for dataset 3) compared to the algorithm using all channels (AC), when CSP is used to extract the features. Furthermore, RCSP could further improve the classification accuracy (81.6% for dataset1, 87.4% for dataset2 and 91.9% for dataset 3), when CCS is used to select the channels. • A new correlation-based channel selection (CCS) method is proposed in this paper. • A novel regularized CSP (RCSP) is proposed to optimize the motor imagery features. • The CCS-RCSP method can significantly improve the classification accuracy. [ABSTRACT FROM AUTHOR]

The start of the cue is often used to initiate the feature window used to control motor imagery (MI)-based brain-computer interface (BCI) systems. However, the time latency during an MI period varies between trials for each participant. Fixing the starting time point of MI features can lead to decreased system performance in MI-based BCI systems. To address this issue, we propose a novel correlation-based time window selection (CTWS) algorithm for MI-based BCIs. Specifically, the optimized reference signals for each class were selected based on correlation analysis and performance evaluation. Furthermore, the starting points of time windows for both training and testing samples were adjusted using correlation analysis. Finally, the feature extraction and classification algorithms were used to calculate the classification accuracy. With two datasets, the results demonstrate that the CTWS algorithm significantly improved the system performance when compared to directly using feature extraction approaches. Importantly, the average improvement in accuracy of the CTWS algorithm on the datasets of healthy participants and stroke patients was 16.72% and 5.24%, respectively when compared to traditional common spatial pattern (CSP) algorithm. In addition, the average accuracy increased 7.36% and 9.29%, respectively when the CTWS was used in conjunction with Sub-Alpha-Beta Log-Det Divergences (Sub-ABLD) algorithm. These findings suggest that the proposed CTWS algorithm holds promise as a general feature extraction approach for MI-based BCIs. [ABSTRACT FROM AUTHOR]

One of the most important issues for the development of a motor-imagery based brain-computer interface (BCI) is how to design a powerful classifier with strong generalization capability. Extreme learning machine (ELM) has recently proven to be comparable or more efficient than support vector machine for many pattern recognition problems. In this study, we propose a multi-kernel ELM (MKELM)-based method for motor imagery electroencephalogram (EEG) classification. The kernel extension of ELM provides an elegant way to circumvent calculation of the hidden layer outputs and inherently encode it in a kernel matrix. We investigate effects of two different kernel functions (i.e., Gaussian kernel and polynomial kernel) on the performance of kernel ELM. The MKELM method is subsequently developed by integrating these two types of kernels with a multi-kernel learning strategy, which can effectively explore the supplementary information from multiple nonlinear feature spaces for more robust classification of EEG. An extensive experimental comparison with two public EEG datasets indicates that the MKELM method gives higher classification accuracy than those of the other competing algorithms. The experimental results confirm that superiority of the proposed MKELM-based method for accurate classification of EEG associated with motor imagery in BCI applications. Our method also provides a promising and generalized solution to investigate the complex and nonlinear information for various applications in the fields of expert and intelligent systems. [ABSTRACT FROM AUTHOR]

Background We analyze the dynamics of rodent EEG amplitude in an experiment accompanied by video recordings. Brain activity of animals is commonly acquired together with a video of behavior, but recordings are rarely combined in analysis. The data acquired is most commonly analyzed separately. To our knowledge, no study has used behavior to improve the analysis of EEG waveforms, specifically for artifact removal – other than through manual editing. Comparison with existing method(s) We explore two approaches: a traditional approach that relies on data preprocessing and artifact rejection by an expert; and an alternative approach that combines analysis of EEG with behavior extracted from video recordings. New method We use the level of activity extracted from the behavioral video as a measure of confidence in the acquired EEG waveform, and as a weighting factor in averaging and statistical comparisons. Results We find in analysis of the EEG that the two approaches lead to similar conclusions, but the analysis leveraging behavioral data achieves this while avoiding many subjective choices often required for artifact rejection and data preprocessing. Conclusions The methods we describe allow for the inclusion of all recorded data in the analysis, thereby making statistical tests more friendly to interpretation, and making the data processing transparent and reproducible. [ABSTRACT FROM AUTHOR]

L1-regularized multiway canonical correlation analysis (L1-MCCA) has been introduced to reference signal optimization in steady-state visual evoked potential (SSVEP)-based brain–computer interface (BCI). The effectiveness of L1-regularization on significant trial selection highly depends on the regularization parameter setting, which can be typically determined by cross-validation (CV). However, CV will substantially reduce the practicability of BCI system due to additional data requirement for the parameter validation and relatively high computational cost. To solve the problem, this study proposes a Bayesian version of L1-MCCA (called SBMCCA) by exploiting sparse Bayesian learning. The SBMCCA method avoids CV and can efficiently estimate the model parameters under the Bayesian evidence framework. Experimental results show that the SBMCCA method achieved comparable recognition accuracy but much higher computational efficiency in contrast to the L1-MCCA method. [ABSTRACT FROM AUTHOR]

Background Multichannels used in brain–computer interface (BCI) systems contain redundant information and cause inconvenience for practical application. Channel selection can enhance the performance of BCI by removing task-irrelevant and redundant channels. Sequential floating forward selection (SFFS) is an intelligent search algorithm and is considered one of the best feature selection methods in the literature. However, SFFS is time consuming when the number of features is large. Method In this study, the SFFS method was improved to select channels for the common spatial pattern (CSP) in motor imagery (MI)-based BCI. Based on the distribution of channels in the cerebral cortex, the adjacent channels would be treated as one feature for selection. Thus, in the search process, the improved SFFS could select or remove several channels in each iteration and reduce the total computation time. Results The improved SFFS yielded significantly better performance than using all channels ( p <0.01) and support vector machine recursive feature elimination method ( p <0.05). The computation time of the proposed method was significantly reduced ( p <0.005) compared with the original SFFS method. Conclusions This study improved the SFFS method to select channels for CSP in MI-based BCI. The improved SFFS method could significantly reduce computation time compared with the original SFFS without compromising the classification accuracy. This study provided a way to optimize electroencephalogram channels, which combined the distribution of channels and the intelligent selection method (SFFS). Improvements were mainly in the perspective of reducing computation time, which leads to convenience in the practical application of BCI systems. [ABSTRACT FROM AUTHOR]

Nonnegative tensor factorization (NTF) has been widely applied in high-dimensional nonnegative tensor data analysis. However, most of the existing algorithms suffer from slow convergence caused by the nonnegativity constraint and hence their practical applications are severely limited. In this study, we propose a new algorithm called FastNTF_APG to speed up NTF by combining accelerated proximal gradient and low-rank approximation. Experimental results demonstrate that FastNTF_APG achieves significantly higher computational efficiency than state-of-the-art NTF algorithms. [ABSTRACT FROM AUTHOR]