Multi-step time series forecasting often suffers from accuracy degradation as the prediction horizon increases, mainly because existing models lack explicit mechanisms to capture and adapt to the evolving dynamics of forecast errors over time. To address this limitation, we propose a dynamic error modification method, termed Adaptive Residual Decay Mechanism (ARDM), which establishes an end-to-end predictive optimization method aimed at enhancing forecasting stability and generalization across diverse temporal patterns and application scenarios. ARDM comprehensively integrates data preprocessing, initial forecasting, residual analysis, error modification, and final output. By constructing residual sequences and extracting their underlying temporal dependencies, ARDM effectively incorporates both short-term and long-term error evolution patterns. Within a symmetrical architecture, a time-sensitive adaptive decay function is introduced to dynamically estimate and adjust horizon-dependent forecast errors. Optimal decay functions and parameters are selected through a multi-metric joint loss function, which balances sensitivity to error magnitude with robustness against directional deviations. Furthermore, ARDM systematically models and exploits historical residual information during the observation phase, enabling stable and horizon-aware refinement of prediction errors through structured residual dependencies. Empirical results on multiple real-world datasets demonstrate that ARDM consistently outperforms mainstream baseline methods across various metrics (MAE, MSE, MAPE, RMSE, SSE, and IA) validating its high accuracy and strong robustness in complex time series forecasting tasks.

Large-scale Internet of Things (IoT) systems are characterized by massive numbers of interconnected devices, heterogeneous dynamics, and complex interaction structures, which can be effectively modeled using complex networks. In many IoT applications, secure cluster synchronization is essential for coordinated and reliable operation, yet it is highly vulnerable to cyber-attacks, particularly replay attacks that maliciously reuse previously transmitted but valid data. This paper investigates the secure cluster synchronization problem for discrete-time nonlinear complex networks representing IoT systems under stochastic replay attacks. A probabilistic replay attack model with bounded consecutive attack duration is introduced to capture the random and intermittent characteristics of realistic attack behaviors. To mitigate the adverse impact of replayed information, a PID-based cluster synchronization control strategy is developed, where proportional, integral, and derivative actions are jointly exploited to enhance robustness against outdated and compromised signals. By constructing an appropriate Lyapunov functional and employing stochastic analysis techniques, sufficient conditions are derived to guarantee asymptotic mean-square cluster synchronization. A systematic controller synthesis procedure is further provided. Numerical simulations demonstrate the effectiveness and improved resilience of the proposed approach compared with conventional proportional control schemes.

This paper is concerned with the asynchronous sampled-data state estimation problems for a class of continuous-time nonlinear complex networks. A novel asynchronous sampled-data estimator is constructed with matrix exponential gains to estimate the states of network nodes, which allows each node to independently sample and transmit the measured signals at its own designated time instants. It is demonstrated that the utilization of matrix exponential gains is capable of enlarging the maximum-allowable bound of the sampling intervals. Moreover, a modified Halanay-type inequality is derived to facilitate the analysis of estimation errors. Accordingly, by leveraging the Lyapunov stability theory, some sufficient conditions are obtained to guarantee the global exponential stability of the estimation error dynamics. In addition, the maximum-allowable bound of the sampling intervals is explicitly characterized by resorting to an algebraic inequality, and a convex optimization method is adopted with the aim of maximizing such an allowable bound. Finally, some numerical simulations are conducted to validate the feasibility and usefulness of the established theoretical results.

In this article, the encoding-decoding-based state estimation problem is investigated for a class of continuous-time nonlinear complex networks (CNs) subject to communication bandwidth constraints. Based on the sampled outputs from a subset of network nodes, a novel dynamic event-driven encoding mechanism is integrated into the design of state estimator, where a time-varying auxiliary parameter is utilized to modulate the triggering condition in a dynamical fashion, enabling the event detector to decide whether the data packet should be released at the periodic sampling instants. Specifically, when the dynamic triggering condition is satisfied, the data are first encoded into a codeword and subsequently transmitted to the estimator through a digital communication channel. The Zeno behavior can be naturally prevented due to the periodic feature of the proposed event detector. By leveraging the Lyapunov theory and the matrix inequality techniques, sufficient conditions are established to ensure the exponential stability of the estimation error system. In addition, a convex optimization approach is employed to design the estimator gain with the goal of maximizing the allowable bound of the sampling intervals. Finally, an illustrative example and a practical example involving a three-area power system are provided to showcase the effectiveness of the proposed state estimation method.

The Real2Sim2Real (R2S2R) paradigm is critical for advancing robotic learning. Existing methods lack a comprehensive solution to accurately reconstruct real-world objects with both spatial representations and their associated physics attributes in the Real2Sim stage. We propose a Real2Sim pipeline to generate digital assets enabling high-fidelity simulation. We design a hybrid repre-sentation model that integrates mesh geometry, 3D Gaussian kernels, and physics attributes to enhance the representation of robotic arms in digital assets. This hybrid representation is implemented through a Gaussian-Mesh-Pixel binding technique, which establishes an isomorphic mapping between mesh vertices and the Gaussian model. This enables a fully differentiable rendering pipeline that can be optimized through numerical solvers, achieves high-fidelity rendering via Gaussian Splatting, and facilitates physically plausible simulation of the robotic arm's interaction with its environment through mesh geometry. With the digital assets, we propose a fully manipulable Real2Sim pipeline that standardizes coordinate systems and scales, ensuring the seamless integration of multiple components. To demonstrate its effectiveness, we include datasets covering various robotic manipulation tasks with their mesh reconstructions. Our model achieves state-of-the-art results in realistic rendering and mesh reconstruction quality for robotic applications. Our code and datasets will be made publicly available at robostudioapp.com.

With origins in game-theory, probabilistic values like Shapley values, Banzhaf values, and semi-values have emerged as a central tool in explainable AI. They are used for feature attribution, data attribution, data valuation, and more. Since all of these values require exponential time to compute exactly, research has focused on efficient approximation methods using two techniques: Monte Carlo sampling and linear regression formulations. In this work, we present a new way of combining both of these techniques. Our approach is more flexible than prior algorithms, allowing for linear regression to be replaced with any function family whose probabilistic values can be computed efficiently. This allows us to harness the accuracy of tree-based models like XGBoost, while still producing unbiased estimates. From experiments across eight datasets, we find that our methods give state-of-the-art performance for estimating probabilistic values. For Shapley values, the error of our methods is up to $6\times$ lower than Permutation SHAP (the most popular Monte Carlo method), $2.75\times$ lower than Kernel SHAP (the most popular linear regression method), and $1.75\times$ lower than Leverage SHAP (the prior state-of-the-art Shapley value estimator). For more general probabilistic values, we can obtain error up to $60\times$ lower than prior work.

Recent advances in language models (LMs) open new opportunities for schema matching (SM). Recent approaches have shown their potential and key limitations: while small LMs (SLMs) require costly, difficult-to-obtain training data, large LMs (LLMs) demand significant computational resources and face context window constraints. We present Magneto, a cost-effective and accurate solution for SM that combines the advantages of SLMs and LLMs to address their limitations. By structuring the SM pipeline in two phases, retrieval and reranking, Magneto can use computationally efficient SLM-based strategies to derive candidate matches which can then be reranked by LLMs, thus making it possible to reduce runtime while improving matching accuracy. We propose (1) a self-supervised approach to fine-tune SLMs which uses LLMs to generate syntactically diverse training data, and (2) prompting strategies that are effective for reranking. We also introduce a new benchmark, developed in collaboration with domain experts, which includes real biomedical datasets and presents new challenges for SM methods. Through a detailed experimental evaluation, using both our new and existing benchmarks, we show that Magneto is scalable and attains high accuracy for datasets from different domains.

This article examines the nonfragile bipartite tracking consensus issue in the context of sampled-data nonlinear multiagent systems (MASs) undergoing denial-of-service (DoS) attacks and control gain fluctuations, where both cooperative and competitive interactions between the agents over the network are taken into account. During the DoS attacks, with communication services being denied, a halt in data transmission among the agents is experienced, which might result in performance degradation, undesirable oscillatory behavior, or even hinders the agents from performing their intended tasks. Consequently, there emerges a pressing requirement for the analysis and design of a secure bipartite tracking consensus protocol for MASs under the threat of DoS attacks. In pursuit of this goal, a modified Halanay-like inequality is initially established, which provides a basis for us to derive certain sufficient conditions ensuring the MASs to achieve the bipartite tracking consensus, despite the disruptive presence of malicious DoS attacks. Additionally, the control gain matrix can be readily computed by resolving a collection of linear matrix inequalities. For specific scenarios that demand reduced computational complexity, the matrix decoupling method is introduced, enabling a reduction in the dimensionality of the matrix inequalities and, consequently, facilitating its straightforward application to large-scale MASs. This article culminates in a numerical simulation, which is performed to validate the developed results.

This article is concerned with the privacy preservation problem for scaled consensus in nonlinear multiagent systems (MASs) through sampled data. In scaled consensus, the aim is for agents’ states to achieve specified proportions rather than converging to a single value, and this approach encompasses standard consensus, bipartite consensus, and cluster consensus within its framework. To prevent the leakage of sensitive data during communication, a novel privacy-preserving scaled distributed protocol is proposed. This protocol uses homomorphic encryption, whereby agents initially encrypt their information and transmit it in ciphertext form to neighboring agents. The control protocol is then reconstructed by the agents using the encrypted information received from their neighbors. A modified Halanay-like inequality is formulated and, by leveraging algebraic graph theory and the Lyapunov stability theorem, sufficient conditions are established to ensure that the MASs can achieve exponentially ultimately bounded scaled consensus. Furthermore, a convex optimization method is adopted to identify the optimal control gain so as to maximize the allowable sampling interval bound. The theoretical results are substantiated through a numerical simulation.

Existing deep-learning approaches to semantic column type annotation (CTA) have important shortcomings: they rely on semantic types which are fixed at training time; require a large number of training samples per type; incur high run-time inference costs; and their performance can degrade when evaluated on novel datasets, even when types remain constant. Large language models have exhibited strong zero-shot classification performance on a wide range of tasks and in this paper we explore their use for CTA. We introduce ArcheType, a simple, practical method for context sampling, prompt serialization, model querying, and label remapping, which enables large language models to solve CTA problems in a fully zero-shot manner. We ablate each component of our method separately, and establish that improvements to context sampling and label remapping provide the most consistent gains. ArcheType establishes a new state-of-the-art performance on zero-shot CTA benchmarks (including three new domain-specific benchmarks which we release along with this paper), and when used in conjunction with classical CTA techniques, it outperforms a SOTA DoDuo model on the fine-tuned SOTAB benchmark.

In the face of the growing global influence and prevalence of Chinese technology companies, governments world-wide have expressed concern and mistrust toward these companies. There is a scarcity of research that specifically examines the widespread public response to this phenomenon on a large scale. This study aims to fill in the gap in understanding online public opinion toward Chinese technology companies using Reddit data, a popular news-oriented social media platform. We employ the state-of-the-art transformer model to build a reliable sentiment classifier. We then use LDA to extract the topics associated with positive and negative comments. We also conduct content analysis by studying the changes in the semantic meaning of the companies’ names over time. Our main findings include the following: 1) Notable difference exists in the proportions of positive comments (8.42%) and negative comments (14.12%); 2) Positive comments are mostly associated with the companies’ consumer products, such as smartphones, laptops, and wearable electronics. Negative comments have a more diverse topic distribution (notable topics include criticism toward the platform, dissatisfaction with the companies’ smartphone products, companies’ ties to the Chinese government, data security concerns, 5G construction, and general political discussions); and 3) Characterization of each technology company is usually centered around a particular predominant theme related to the company, while real-world political events may trigger drastic changes in users’ characterization.

The main purpose of this article is to investigate the consensus of linear multiagent networks with time-varying characteristics under sampled-data communications, where the time-varying characteristics include both time-varying topologies and the node's linear time-varying dynamics. By using the decoupling method, we prove that the sampled-data consensus problem of multiagent networks is equal to the stability problem of sampled-data systems. Then, the globally asymptotical consensus is investigated for multiagent networks with time-varying characteristics by virtue of the Lyapunov function method. It should be noted that when the Lyapunov function method is utilized to investigate the stability problem of control systems, it is always assumed that the derivative of the constructed Lyapunov function is not more than zero. This assumption is removed here and as a replacement, the average value of the derivative of the Lyapunov function in a period to be negative is needed.

In this paper, the sampled-based consensus problem is investigated for a class of nonlinear multiagent systems subjected to deception attacks. Due to network fluctuations and limited resources, deception attacks might destroy the sampled-data in communication networks. Additionally, the success of deception attacks greatly depends on some randomly fluctuated factors. This paper takes into account the deception attacks that are randomly launched at each sampling instant. The eigenvector of Laplacian matrix is utilized to construct a novel Lyapunov functional. Then, the decoupled criterion connected with eigenvalues of Laplacian matrix is obtained such that the addressed multiagent systems achieve the mean square consensus. Furthermore, the solutions of a set of matrix inequalities represent the controller gain matrix. Finally, a numerical example is provided to validate the effectiveness of the derived results.

In this paper, the quasi-consensus problem is investigated for a class of heterogeneous-switched nonlinear multiagent systems, in which both cooperation and competition interactions are considered simultaneously. By means of the Lyapunov function method, we show that quasi-consensus can be ensured for switched multiagent systems under the assumption that the activation time of cooperation interactions is sufficiently large. Moreover, a new Lyapunov function is considered to provide the lower and upper bounds of switching intervals explicitly. Thus, these bounds can be used to obtain less conservative stability results of switched systems. Furthermore, the established results are specialized to both the traditional consensus case and the stability of linear-switched systems. Finally, simulations are given to illustrate the theoretical results derived in this paper.

The phenomena of unknown correlations are ubiquitously existing in general distributed filtering problems over sensor networks. And the covariance intersection (CI) fusion rule is an effective tool to tackle with this phenomena. During the recent years, the related CI-based Kalman consensus filters (CIKCFs) have been adopted to deal with unknown correlations in sensor networks. However, a systematic stability analysis result for the general CIKCF in the time-varying system setting remains to be established. This paper is written for this purpose. First, a general CIKCF with full features of CI is presented. Accordingly, the conditions for CIKCF to reach consensus with varying weights are investigated. Furthermore, a novel detectability condition, i.e., collectively uniform detectability, is proposed to ensure the error covariances of the CIKCF are uniformly bounded. Based on this condition, the estimation errors are further proven to be exponentially bounded in mean square with the aid of the stochastic stability lemma. Finally, an example is given to validate the effectiveness of the theoretical results.

This paper is concerned with the distributed H∞state estimation for a discrete-time target linear system over a filtering network with time-varying and switching topology and partial information exchange. Both filtering network topology switching and partial information exchange between filters are simultaneously considered in the filter design. The topology under consideration evolves not only over time but also by an event switch which is assumed to be subject to a nonhomogeneous Markov chain. The probability transition matrix of the nonhomogeneous Markov chain is time-varying. In the filter information exchange, partial state estimation information and channel noise are simultaneously considered. In order to design such a switching filtering network with partial information exchange, stochastic Markov stability theory is developed. The switching topology-dependent filters are derived to guarantee an optimal H∞disturbance rejection attenuation level for the estimation disagreement of the filtering network. It is shown that the addressed H∞state estimation problem is turned into a switching topology-dependent optimal problem. The distributed filtering problem with complete information exchanges from its neighbors is also investigated. An illustrative example is given to show the applicability of the obtained results.

In this paper, the state estimation problem is investigated for a class of continuous-time complex networks with bounded distributed delay. For the network under consideration, only the outputs from a fraction of nodes are available and this put forward the new challenge, that is, the so-called partial-nodes-based state estimation problem. In order to reduce the usage of the communication resources, a general event-triggering rule is considered in the design of the estimator. A novel estimator is constructed and, by constructing novel Lyapunov-Krasovskii functionals, some easy-to-check conditions are derived such that the error dynamics is exponentially ultimately bounded. Furthermore, it is shown that the Zeno behavior can be excluded from the event-triggering rules. Numerical simulations are presented to further illustrate the effectiveness of the theoretical results.

This paper addresses the problem of stability for continuous-time and discrete-time quaternion-valued neural networks (QVNNs) with linear threshold neurons. Applying the semidiscretization technique to the continuous-time QVNNs, the discrete-time analogs are obtained, which preserve the dynamical characteristics of their continuous-time counterparts. Via the plural decomposition method of quaternion, homeomorphic mapping theorem, as well as Lyapunov theorem, some sufficient conditions on the existence, uniqueness, and global asymptotical stability of the equilibrium point are derived for the continuous-time QVNNs and their discrete-time analogs, respectively. Furthermore, a uniform sufficient condition on the existence, uniqueness, and global asymptotical stability of the equilibrium point is obtained for both continuous-time QVNNs and their discrete-time version. Finally, two numerical examples are provided to substantiate the effectiveness of the proposed results.

In this brief, the new problem of partial-nodes-based (PNB) state estimation problem is investigated for a class of complex network with unbounded distributed delays and energy-bounded measurement noises. The main novelty lies in that the states of the complex network are estimated through measurement outputs of a fraction of the network nodes. Such fraction of the nodes is determined by either the practical availability or the computational necessity. The PNB state estimator is designed such that the error dynamics of the network state estimation is exponentially ultimately bounded in the presence of measurement errors. Sufficient conditions are established to ensure the existence of the PNB state estimators and then the explicit expression of the gain matrices of such estimators is characterized. When the network measurements are free of noises, the main results specialize to the case of exponential stability for error dynamics. Numerical examples are presented to verify the theoretical results.

In this correspondence paper, we provide a new look at the boundedness problems of error covariance of Kalman filtering. First, by utilizing the mathematical induction technique, a new bound function which is dependent on system parameters is proposed. In this manner, the boundedness problems of the error covariance can be converted to the study of the corresponding uniform bounds of the bound function. Second, based on such a bound function, the dynamic behaviors, monotonicities, and boundedness problems of error covariance are deeply explored. Consequently, a few quantitative results under minimal conditions about the uniform bounds on error covariance are obtained. Finally, examples are given to verify the correctness and effectiveness of our theoretical analyses.

In this paper, we are devoted to investigate the consensus-based distributed state estimation problems for a class of sensor networks within the unscented Kalman filter (UKF) framework. The communication status among sensors is represented by a connected undirected graph. Moreover, a weighted average consensus-based UKF algorithm is developed for the purpose of estimating the true state of interest, and its estimation error is bounded in mean square which has been proven in the following section. Finally, the effectiveness of the proposed consensus-based UKF algorithm is validated through a simulation example.

This paper is concerned with optimal communication network-based H∞ quantized control for a discrete-time neural network with distributed time delay. Control of the neural network (plant) is implemented via a communication network. Both quantization and communication network-induced data packet dropouts are considered simultaneously. It is assumed that the plant state signal is quantized by a logarithmic quantizer before transmission, and communication network-induced packet dropouts can be described by a Bernoulli distributed white sequence. A new approach is developed such that controller design can be reduced to the feasibility of linear matrix inequalities, and a desired optimal control gain can be derived in an explicit expression. It is worth pointing out that some new techniques based on a new sector-like expression of quantization errors, and the singular value decomposition of a matrix are developed and employed in the derivation of main results. An illustrative example is presented to show the effectiveness of the obtained results.

In this paper, the synchronization problem is studied for an array of N identical delayed neutral-type neural networks with Markovian jumping parameters. The coupled networks involve both the mode-dependent discrete-time delays and the mode-dependent unbounded distributed time delays. All the network parameters including the coupling matrix are also dependent on the Markovian jumping mode. By introducing novel Lyapunov-Krasovskii functionals and using some analytical techniques, sufficient conditions are derived to guarantee that the coupled networks are asymptotically synchronized in mean square. The derived sufficient conditions are closely related with the discrete-time delays, the distributed time delays, the mode transition probability, and the coupling structure of the networks. The obtained criteria are given in terms of matrix inequalities that can be efficiently solved by employing the semidefinite program method. Numerical simulations are presented to further demonstrate the effectiveness of the proposed approach.

In this paper, the problem of stochastic synchronization analysis is investigated for a new array of coupled discrete-time stochastic complex networks with randomly occurred nonlinearities (RONs) and time delays. The discrete-time complex networks under consideration are subject to: 1) stochastic nonlinearities that occur according to the Bernoulli distributed white noise sequences; 2) stochastic disturbances that enter the coupling term, the delayed coupling term as well as the overall network; and 3) time delays that include both the discrete and distributed ones. Note that the newly introduced RONs and the multiple stochastic disturbances can better reflect the dynamical behaviors of coupled complex networks whose information transmission process is affected by a noisy environment (e.g., internet-based control systems). By constructing a novel Lyapunov-like matrix functional, the idea of delay fractioning is applied to deal with the addressed synchronization analysis problem. By employing a combination of the linear matrix inequality (LMI) techniques, the free-weighting matrix method and stochastic analysis theories, several delay-dependent sufficient conditions are obtained which ensure the asymptotic synchronization in the mean square sense for the discrete-time stochastic complex networks with time delays. The criteria derived are characterized in terms of LMIs whose solution can be solved by utilizing the standard numerical software. A simulation example is presented to show the effectiveness and applicability of the proposed results.

In this paper, the extended Kalman filter (EKF) algorithm is applied to model the gene regulatory network from gene time series data. The gene regulatory network is considered as a nonlinear dynamic stochastic model that consists of the gene measurement equation and the gene regulation equation. After specifying the model structure, we apply the EKF algorithm for identifying both the model parameters and the actual value of gene expression levels. It is shown that the EKF algorithm is an online estimation algorithm that can identify a large number of parameters (including parameters of nonlinear functions) through iterative procedure by using a small number of observations. Four real-world gene expression data sets are employed to demonstrate the effectiveness of the EKF algorithm, and the obtained models are evaluated from the viewpoint of bioinformatics.

In this paper, we introduce a new class of discrete-time neural networks (DNNs) with Markovian jumping parameters as well as mode-dependent mixed time delays (both discrete and distributed time delays). Specifically, the parameters of the DNNs are subject to the switching from one to another at different times according to a Markov chain, and the mixed time delays consist of both discrete and distributed delays that are dependent on the Markovian jumping mode. We first deal with the stability analysis problem of the addressed neural networks. A special inequality is developed to account for the mixed time delays in the discrete-time setting, and a novel Lyapunov-Krasovskii functional is put forward to reflect the mode-dependent time delays. Sufficient conditions are established in terms of linear matrix inequalities (LMIs) that guarantee the stochastic stability. We then turn to the synchronization problem among an array of identical coupled Markovian jumping neural networks with mixed mode-dependent time delays. By utilizing the Lyapunov stability theory and the Kronecker product, it is shown that the addressed synchronization problem is solvable if several LMIs are feasible. Hence, different from the commonly used matrix norm theories (such as the M-matrix method), a unified LMI approach is developed to solve the stability analysis and synchronization problems of the class of neural networks under investigation, where the LMIs can be easily solved by using the available Matlab LMI toolbox. Two numerical examples are presented to illustrate the usefulness and effectiveness of the main results obtained.

This paper is concerned with the robust synchronization problem for an array of coupled stochastic discrete-time neural networks with time-varying delay. The individual neural network is subject to parameter uncertainty, stochastic disturbance, and time-varying delay, where the norm-bounded parameter uncertainties exist in both the state and weight matrices, the stochastic disturbance is in the form of a scalar Wiener process, and the time delay enters into the activation function. For the array of coupled neural networks, the constant coupling and delayed coupling are simultaneously considered. We aim to establish easy-to-verify conditions under which the addressed neural networks are synchronized. By using the Kronecker product as an effective tool, a linear matrix inequality (LMI) approach is developed to derive several sufficient criteria ensuring the coupled delayed neural networks to be globally, robustly, exponentially synchronized in the mean square. The LMI-based conditions obtained are dependent not only on the lower bound but also on the upper bound of the time-varying delay, and can be solved efficiently via the Matlab LMI Toolbox. Two numerical examples are given to demonstrate the usefulness of the proposed synchronization scheme.

In this paper, the synchronization control problem is considered for two coupled discrete-time complex networks with time delays. The network under investigation is quite general to reflect the reality, where the state delays are allowed to be time varying with given lower and upper bounds, and the stochastic disturbances are assumed to be Brownian motions that affect not only the network coupling but also the overall networks. By utilizing the Lyapunov functional method combined with linear matrix inequality (LMI) techniques, we obtain several sufficient delay-dependent conditions that ensure the coupled networks to be globally exponentially synchronized in the mean square. A control law is designed to synchronize the addressed coupled complex networks in terms of certain LMIs that can be readily solved using the Matlab LMI toolbox. Two numerical examples are presented to show the validity of our theoretical analysis results.

In this paper, a synchronization problem is investigated for an array of coupled complex discrete-time networks with the simultaneous presence of both the discrete and distributed time delays. The complex networks addressed which include neural and social networks as special cases are quite general. Rather than the commonly used Lipschitz-type function, a more general sector-like nonlinear function is employed to describe the nonlinearities existing in the network. The distributed infinite time delays in the discrete-time domain are first defined. By utilizing a novel Lyapunov-Krasovskii functional and the Kronecker product, it is shown that the addressed discrete-time complex network with distributed delays is synchronized if certain linear matrix inequalities (LMIs) are feasible. The state estimation problem is then studied for the same complex network, where the purpose is to design a state estimator to estimate the network states through available output measurements such that, for all admissible discrete and distributed delays, the dynamics of the estimation error is guaranteed to be globally asymptotically stable. Again, an LMI approach is developed for the state estimation problem. Two simulation examples are provided to show the usefulness of the proposed global synchronization and state estimation conditions. It is worth pointing out that our main results are valid even if the nominal subsystems within the network are unstable.

In this letter, the global asymptotic stability analysis problem is considered for a class of stochastic Cohen-Grossberg neural networks with mixed time delays, which consist of both the discrete and distributed time delays. Based on an Lyapunov-Krasovskii functional and the stochastic stability analysis theory, a linear matrix inequality (LMI) approach is developed to derive several sufficient conditions guaranteeing the global asymptotic convergence of the equilibrium point in the mean square. It is shown that the addressed stochastic Cohen-Grossberg neural networks with mixed delays are globally asymptotically stable in the mean square if two LMIs are feasible, where the feasibility of LMIs can be readily checked by the Matlab LMI toolbox. It is also pointed out that the main results comprise some existing results as special cases. A numerical example is given to demonstrate the usefulness of the proposed global stability criteria.