Federated Learning inherently faces the challenge of balancing privacy protection and classification accuracy due to the risks associated with parameter sharing and the limitations of model performance. This paper proposes Fed-WGCA, a novel federated learning framework that integrates Wasserstein GAN with Gradient Penalty (WGAN-GP) and Coordinate Attention mechanisms. WGAN-GP enhances data privacy by generating high-quality synthetic data, mitigates the impact of imbalanced samples, and significantly improves model stability and performance. The Coordinate Attention mechanism optimizes feature extraction by fusing spatial and channel information. Experimental results demonstrate that the proposed framework Fed-WGCA separately achieves accuracy improvements of 5.99% and 6.12% on the Fashion-MNIST and CIFAR-10 datasets compared with state-of-the-art best results while exhibiting enhanced robustness against membership inference, image reconstruction attacks and poisoning attack.

Influence maximization (IM) algorithms aim to iteratively identify a seed set that could maximize the spreading range. In this paper, we concentrate on the hypergraph influence maximization (HyperIM) problem. The overlapping neighborhood caused by the higher-order interactions leads to an overestimation of the diffusion capability of candidate nodes. Moreover, selecting the candidate node with a high infected probability as a new seed node is low payoff with a small influence range gain. Thus, we develop adaptive metrics and propose two algorithms, i.e., high adaptive contact efficiency (HACE) algorithm and high contact with a low infected probability (HCLI) algorithm. First, we penalize the contribution of the neighborhood of the seed set to the evaluation of the influence gain to the seed set to correct the impact of overlapping influence. Additionally, the proposed HACE algorithm uses the being contacted capability to reveal the infected possibility of candidate nodes, while the proposed HCLI algorithm estimates the global infected probability of nodes. The experiments and analysis on eight real-world hypergraphs demonstrate the better balance of the HACE and HCLI algorithms than the state-of-the-art (SOTA) algorithms in selecting influential seed set and ensuring computational efficiency. Compared with the existing SOTA algorithms, HACE and HCLI run at least ten times faster than SOTA, and at most nearly 70 times faster. On large-scale hypergraphs, the HACE and HCLI algorithms still show great computational efficiency and significantly improved performances compared with other low-time complexity algorithms.

Previous deep image registration methods that employ single homography, multi-grid homography, or thin-plate spline often struggle with real scenes containing depth disparities due to their inherent limitations. To address this, we propose an Exponential-Decay Free-Form Deformation Network (EDFFDNet), which employs free-form deformation with an exponential-decay basis function. This design achieves higher efficiency and performs well in scenes with depth disparities, benefiting from its inherent locality. We also introduce an Adaptive Sparse Motion Aggregator (ASMA), which replaces the MLP motion aggregator used in previous methods. By transforming dense interactions into sparse ones, ASMA reduces parameters and improves accuracy. Additionally, we propose a progressive correlation refinement strategy that leverages global-local correlation patterns for coarse-to-fine motion estimation, further enhancing efficiency and accuracy. Experiments demonstrate that EDFFDNet reduces parameters, memory, and total runtime by 70.5%, 32.6%, and 33.7%, respectively, while achieving a 0.5 dB PSNR gain over the state-of-the-art method. With an additional local refinement stage,EDFFDNet-2 further improves PSNR by 1.06 dB while maintaining lower computational costs. Our method also demonstrates strong generalization ability across datasets, outperforming previous deep learning methods.

The prediction of future dynamics on networks is a challenge. Unfortunately, due to the noise in the sampling process, or the low resolution of observational data, it is hardly feasible to get the complete topology of real-world networks. Moreover, the higher-order interactions among nodes add the difficulty to accurate prediction of dynamics on networks. In this work, we proposed a two-step approach for higher-order dynamics prediction without network topology. First, the observations of nodal dynamics of a specific dynamical model on unknown hypergraphs are collected to solve an optimization problem by Ridge regression, to obtain a surrogate incidence matrix. Second, the prediction is obtained by iterating the equation of the dynamical model with the surrogate incidence matrix. We define the average relative prediction error to evaluate the performance of our prediction method, and a wide range of hypergraph dynamics with different parameters are predicted. The prediction accuracy is positively correlated with the number of hyperedges the hypergraph contains and the contact rate in the dynamical model.

At present, hyperspectral image classification (HSIC) technology has been warmly concerned in all walks of life, especially in agriculture. However, existing classification methods operate under the closed-set assumption, which deviates from the real world with open properties. At the same time, there are more serious phenomena of different crops with similar spectrum and same crops with different spectrum in agricultural hyperspectral data, which is also a great challenge to existing methods. In this work, a fractional Fourier based frequency-spatial-spectral prototype network is proposed to address the challenges of open-set hyperspectral image classification in agricultural scenarios. Firstly, fractional Fourier transform is introduced into the network to combine the information in the frequency domain with the spatial-spectral information, so as to expand the difference between different classes on the premise of ensuring the similarity between classes. Then, the prototype learning strategy is introduced into the network to improve the feature recognition capability of the network through prototype loss. Finally, in order to break the stubbornly closed-set property of closed-set classification method, the open-set recognition module is proposed. The difference between the prototype vector and the feature vector is used to judge the unknown class. Experiments on three agricultural hyperspectral datasets show that this method can effectively identify unknown class without sacrificing the classification accuracy of closed-set, and has satisfactory classification performance.

The increasing prevalence of encrypted communication on the modern internet has presented new challenges for traffic classification and network management. Traditional traffic classification methods cannot handle encrypted traffic effectively. Meanwhile, many existing methods either rely on hand-crafted features or fail to extract the underlying interaction patterns between data packets adequately. In this paper, we propose a novel encrypted traffic classification method called the Attention-based Vision Transformer and Spatiotemporal for Traffic Classification (ATVITSC). In the preprocessing stage, packet-level images within a session, generated from the payload of data packets, are combined into a session image to mitigate information confusion. In the classification stage, session images are first processed by the packet vision transformer (PVT) module, which employs the transformer encoder and multi-head self-attention mechanism, to capture the global features. In parallel, session images are also processed by the spatiotemporal feature extraction (STFE) module, where spatial features of packets are extracted by the convolution operation with the attention mechanism and temporal features between packets are then combined by the bidirectional Long Short-Term Memory (LSTM). The global and spatiotemporal features are fused in the feature fusion classification (FFC) module by a dynamic weighting mechanism and encrypted traffic is finally classified based on the fused features. Comprehensive experiments on various types of encrypted traffic, including virtual private network (VPN), onion router (Tor), malicious traffic, and mobile traffic, show that the ATVITSC successfully improves the macro-f1 scores to 97.88%, 98.79%, 99.67%, 94.90%, respectively. The results also reveal that the ATVITSC exhibits better classification performance and generalization ability than the state-of-the-art methods.

Edge networks are providing services for an increasing number of companies, and they can be used for communication between edge devices and edge gateways. However, the performance of edge devices varies greatly, and it is not easy to upgrade low-performance edge devices. Therefore, cyber attackers can use the vulnerability of edge devices to implement advanced persistent threat attacks. This article proposes a network verification framework for edge networks that can minimize the upgrades needed to strengthen edge network security. First, the communication parties use the data transmitted by the given edge network. Our method uses our proposed PacketVerifier to attach verification information to the packet after it is sent and to verify and restore the packet before it reaches the receiver. Second, due to the performance requirements of edge networks, we design a new data processing structure, namely, a sliding window double ring, to improve the performance of strict sequential protocols in parallel validation. Finally, experimental simulations show that our parallel processing algorithm has good performance in terms of network bandwidth compared with two existing packet processing algorithms. Furthermore, the proposed packet with verification information is compatible with the existing network topology, which helps PacketVerifier establish trustworthy transmission in a zero-trust environment.

Incipient faults are difficult to be detected due to the intrinsic fault tolerance of traditional controller, but it should be eliminated as soon as possible before it deteriorates with time into something more serious. As a consequence of an intrinsic inability to assess whether a fault occurs based on output residual, the existing detection methods are failure for incipient fault. So the aim of active fault detection (AFD) is to make the system be unstable when incipient fault has occurred, which drives rapid fault detection. The fault-tolerant control (FTC) is designed to maintain the system stable and eliminate the fault impact without shutting the process down even if faults occur. In this paper, the Youla-Jabr-Bongiorno-Kucera (YJBK) parameter is employed to build the AFD and the FTC based on the relationship analysis between the fault and the dual YJBK parameter. A new structure of the tolerant controller parameter for FTC is designed, named as partial active FTC (PAFTC). PAFTC is dependent upon the fault detection information but not the fault size considering the parameter fault with unknown size and known form. A unified operation architecture for AFD and PAFTC with different YJBK parameters for incipient faults is proposed. Some illustrative examples are given to indicate the effectiveness of the proposed unified operation architecture.

We study the problem of question topic classification using a very large real‐worldCommunityQuestionAnswering (CQA) dataset fromYahoo!Answers. The dataset comprises 3.9 million questions and these questions are organized into more than 1,000 categories in a hierarchy. To the best knowledge, this is the first systematic evaluation of the performance of different classification methods on question topic classification as well as short texts. Specifically, we empirically evaluate the following in classifying questions intoCQAcategories: (a) the usefulness of n‐gram features and bag‐of‐word features; (b) the performance of three standard classification algorithms (naive Bayes, maximum entropy, and support vector machines); (c) the performance of the state‐of‐the‐art hierarchical classification algorithms; (d) the effect of training data size on performance; and (e) the effectiveness of the different components ofCQAdata, including subject, content, asker, and the best answer. The experimental results show what aspects are important for question topic classification in terms of both effectiveness and efficiency. We believe that the experimental findings from this study will be useful in real‐world classification problems.

Test case prioritization is an effective and practical technique of regression testing. To illustrate its effectiveness, many test metrics were proposed. In this paper, the physical meanings of these metrics were explained and their limitations were pointed out. Then, an improved metric and its extension for test case prioritization were proposed. The case study indicates that, compared with existing metrics, our new metric can provide much more precise illustration of the effectiveness of test case prioritization techniques.

Test case prioritization is an effective and practical technique that helps to increase the rate of regression fault detection when software evolves. Numerous techniques have been reported in the literature on prioritizing test cases for regression testing. However, existing prioritization techniques implicitly assume that source or binary code is available when regression testing is performed, and therefore cannot be implemented when there is no program source or binary code to be analyzed. In this paper, we presented a new technique for black box regression testing, and we performed an experiment to measure our technique. Our results show that the new technique is helpful to improve the effectiveness of fault detection when performing regression test in black box environment.