Showing papers in "Neural Networks in 2017"

TL;DR: It is proved that deep ReLU networks more efficiently approximate smooth functions than shallow networks and adaptive depth-6 network architectures more efficient than the standard shallow architecture are described.

TL;DR: A frame-based formulation to SER is described that relies on minimal speech processing and end-to-end deep learning to model intra-utterance dynamics and is used to empirically explore feed-forward and recurrent neural network architectures and their variants.

TL;DR: It is shown by theoretic proof that the estimation bound of the settling time given in this paper is less conservative and more accurate compared with the classical results.

TL;DR: A novel impulse pinning strategy involving pinning ratio is proposed and a general criterion is derived to ensure an array of neural networks with two different topologies synchronizes with the desired trajectory.

TL;DR: A multi-manifold regularized non-negative matrix factorization framework (MMNMF) which can preserve the locally geometrical structure of the manifolds for multi-view clustering.

TL;DR: Considering the spatial relation of a pixel to its neighborhood, a new deep patch-based CNN system tailored for medium-resolution remote sensing data is proposed and it is believed that much more accurate land cover datasets can be produced over large areas.

TL;DR: A stacked ELM architecture in the CNN framework is proposed using an extreme learning machine (ELM) and the backpropagation algorithm is modified to find the targets of hidden layers and effectively learn network weights while maintaining performance.

TL;DR: A flexible Self-Taught Convolutional neural network framework for Short Text Clustering, which can flexibly and successfully incorporate more useful semantic features and learn non-biased deep text representation in an unsupervised manner is proposed.

TL;DR: To accomplish the target of fixed-time synchronization, a novel feedback control procedure is designed for the slave neural networks by means of the Filippov discontinuity theories and Lyapunov stability theories, and an upper bound of the settling time is acquired.

TL;DR: Several sufficient conditions are obtained for checking the global robust stability of QVNNs without leakage delay as well as complex-valued neural networks (CVNNs) with both leakage and discrete delays.

TL;DR: A self-organizing neural architecture for incrementally learning to classify human actions from video sequences using a set of hierarchically arranged recurrent networks for the unsupervised learning of action representations with increasingly large spatiotemporal receptive fields is proposed.

TL;DR: The problem of global dissipativity analysis for quaternion-valued neural networks (QVNNs) with time-varying delays is firstly investigated and the positive invariant sets, globally attractive sets and globally exponentially attractive sets are figured out.

TL;DR: The Caputo derivative is employed to evaluate the fractional-order gradient of the error defined as the traditional quadratic energy function and the monotonicity and weak convergence of the proposed approach are proved in detail.

TL;DR: A new Lyapunov functional is constructed to directly propose the asymptotic stability of the inertial neural networks, and some new stability criteria are derived by means of Barbalat Lemma.

TL;DR: By constructing appropriate Lyapunov-like functionals and using the average dwell time technique, some sufficient criteria are derived to guarantee the finite-time synchronization of considered uncertain coupled switched neural networks under asynchronous switching.

TL;DR: The problem of global dissipativity for memristor-based inertial networks with time-varying delay of neutral type is investigated and the globally exponentially attractive sets and positive invariant sets are presented here.

TL;DR: It is proved that SNP-MC systems are Turing universal as both number generating and number accepting devices.

TL;DR: A new representation learning framework called Recommendation via Dual-Autoencoder (ReDa) is proposed, which simultaneously learns the new hidden representations of users and items using autoencoders, and develops a gradient descent method to learn hidden representations.

TL;DR: The results extend and improve the exponential impulsive control theory of neural network system with time-varying delays and reaction-diffusion terms to the fractional-order case and provide a conceptually straightforward mathematical representation of rather complex processes.

TL;DR: A class of fractional-order complex-valued neural networks (FCVNNs) with time delay is established, and some sufficient conditions are derived in order to ensure the global asymptotic stability of the addressed FCVNN’s with time Delay.

TL;DR: Based on the two controllers and two lemmas, the error system is proved to be globally asymptotically stable and even fixed-time stable and some sufficient and easy checked conditions are derived to guarantee the global synchronization of drive and response systems in fixed time.

TL;DR: This paper investigates Mittag-Leffler stability of a class of fractional-order neural networks in the presence of generalized piecewise constant arguments and proves that the existence and uniqueness of the solution of the network holds when some conditions are satisfied.

TL;DR: This paper focuses on command-filtered backstepping adaptive control for a class of strict-feedback nonlinear systems with functional uncertainties, where an NN composite learning technique is proposed to guarantee convergence of NN weights to their ideal values without the PE condition.

TL;DR: New testable algebraic criteria for ensuring the existence and global asymptotic stability of the system equilibrium point are obtained by employing the Kakutani's fixed point theorem of set-valued maps, the comparison theorem, and the stability criterion for FO linear systems with multiple delays.

TL;DR: This new control scheme makes full use of received information and overcomes the shortcomings of mode-dependent and mode-independent control schemes, and has great flexibility in practical applications.

TL;DR: This paper investigates master-slave exponential synchronization for a class of complex-valued memristor-based neural networks with time-varying delays via discontinuous impulsive control via linear matrix inequality (LMI).

TL;DR: A new one-step-ahead model is developed in this paper which combines empirical mode decomposition (EMD) with stochastic time strength neural network (STNN) and the empirical results show that the proposed model indeed displays a good performance in forecasting stock market fluctuations.

TL;DR: Experimental results prove the dominance of HG AR-PNN classifier over the existing classifiers with respect to the stability and improved detection rate for less frequent attacks.

TL;DR: This paper designs two similar feedback controllers for fixed-time synchronization of memristor-based BAM neural networks (MBAMNNs) with discrete delay based on Lyapunov stability theories.

TL;DR: Electroencephalogram waves of user and corresponding global textual comments of the video to understand the user's preference more precisely are fused to predict rating of video-advertisements based on a multimodal framework combining physiological analysis of the user and global sentiment-rating available on the internet.