Search

Network embedding aims to learn a low-dimensional vector for each node in networks, which is effective in a variety of applications such as network reconstruction and community detection. However, the majority of the existing network embedding methods merely exploit the network structure and ignore the rich node attributes, which tend to generate sub-optimal network representation. To learn more desired network representation, diverse information of networks should be exploited. In this paper, we develop a novel deep autoencoder framework to fuse topological structure and node attributes named FSADA. We firstly design a multi-layer autoencoder which consists of multiple non-linear functions to capture and preserve the highly non-linear network structure and node attribute information. Particularly, we adopt a pre-processing procedure to pre-process the original information, which can better facilitate to extract the intrinsic correlations between topological structure and node attributes. In addition, we design an enhancement module that combines topology and node attribute similarity to construct pairwise constraints on nodes, and then a graph regularization is introduced into the framework to enhance the representation in the latent space. Our extensive experimental evaluations demonstrate the superior performance of the proposed method.

Deep neural networks are vulnerable to adversarial examples. However, existing attacks exhibit relatively low efficacy in generating transferable adversarial examples. Improved transferability to address this issue is proposed via a rotation‐invariant attack method that maximizes the loss function w.r.t the random rotated image instead of the original input at each iteration, thus mitigating the high correlation between the adversarial examples and the source models and making the adversarial examples more transferable. Extensive experiments show that the proposed method can significantly improve the transferability of the adversarial examples with almost no extra computational cost and can be integrated into various methods. In addition, when this method is easily applied through a plug‐in, the average attack success rate against six robustly trained models increases by 5.4% over the state‐of‐the‐art baseline method, demonstrating its effectiveness and efficiency. The codes used are publicly available at https://github.com/YeXinD/Rotation‐Invariant‐Attack.

This paper proposes a network anomaly detection model of direct batch growing hierarchical self-organizing mapping based on entropy, which facilitates clear topology representation for the asymmetrically-distributed data. Since the entropy-defined parameters dynamically vary with the incident dataset, that is, follow a data-adaptive manner, the proposed model is naturally valid in all cases with various data types. For fine-grained data distinguishing, a resemble entropy parameter is proposed for the first time to our best knowledge. The experimental results validate that the proposed model achieves a more efficient network anomaly detection than the conventional models, especially for real-world applications with unexpected anomaly data updating.

Multi-view clustering aims to obtain the perfect clusters with a set of feature sets. Many methods learn a common agreement among views to achieve this. However, they may fail to use the specific unique feature of each view, which is important to describe the intrinsic characteristics of data. Unlike many works, we treat each view as a subset feature of data and fuse representations from those unique views to learn an integrated graph for clustering. We propose a novel representations fusion method for MVC. In this method, a regularized semi-nonnegative matrix factorization is proposed to learn the low-dimensional representation of each view, in which, a regularization term is designed to keep the neighboring structure in new low-dimensional space, and it is further updated and utilized to guide the learning process. Then the learned representations is fused to get an integrated graph for describing the image data. Finally, extensive experiments on several real image datasets show that our method achieves better performance than the state-of-the-art methods.

Momentum methods, including heavy-ball~(HB) and Nesterov's accelerated gradient~(NAG), are widely used in training neural networks for their fast convergence. However, there is a lack of theoretical guarantees for their convergence and acceleration since the optimization landscape of the neural network is non-convex. Nowadays, some works make progress towards understanding the convergence of momentum methods in an over-parameterized regime, where the number of the parameters exceeds that of the training instances. Nonetheless, current results mainly focus on the two-layer neural network, which are far from explaining the remarkable success of the momentum methods in training deep neural networks. Motivated by this, we investigate the convergence of NAG with constant learning rate and momentum parameter in training two architectures of deep linear networks: deep fully-connected linear neural networks and deep linear ResNets. Based on the over-parameterization regime, we first analyze the residual dynamics induced by the training trajectory of NAG for a deep fully-connected linear neural network under the random Gaussian initialization. Our results show that NAG can converge to the global minimum at a $(1 - \mathcal{O}(1/\sqrt{\kappa}))^t$ rate, where $t$ is the iteration number and $\kappa > 1$ is a constant depending on the condition number of the feature matrix. Compared to the $(1 - \mathcal{O}(1/{\kappa}))^t$ rate of GD, NAG achieves an acceleration over GD. To the best of our knowledge, this is the first theoretical guarantee for the convergence of NAG to the global minimum in training deep neural networks. Furthermore, we extend our analysis to deep linear ResNets and derive a similar convergence result., Comment: 34 pages

Multi-scale feature fusion has been proven effective in substantial person re-identification (ReID) works. However, the existing multi-scale feature fusion is based on features of different semantic levels. We propose a novel multi-scale and multi-branch feature representation for person ReID, namely Ms-Mb. It merges the features of the same semantic level and integrates attention modules to learn robust and representative feature representations. Through the heterogeneous losses supervision, the final feature representation of the image is more discriminative for person ReID. Sufficient ablation study has proven that the multi-scale feature fusion, the attention module and heterogeneous losses training strategy contribute to the performance boost of Ms-Mb. We have conducted experiments on four mainstream benchmarks including Market1501, DukeMTMC-reID, CUHK03 and MSMT17. Extensive experimental results show that our approach Ms-Mb achieves state-of-the-art performances on Market1501 (Rank-1 = 95.8 % , mAP = 89.9 % ), DukeMTMC-reID (Rank-1 = 90.8 % , mAP = 82.2 % ) and MSMT17 (Rank-1 = 81.9 % , mAP = 59.3 % ) without using additional external data or re-ranking. Our approach yields competitive results compared to the state-of-the-art method on CUHK03 (Rank-1 = 75.4 % , mAP = 72.9 % ) and surpasses the other methods by a large margin.

Applying an accurate parametric prediction model to identify abnormal or false pressurizer water levels (PWLs) is critical to the safe operation of marine pressurized water reactors (PWRs). Recently, deep-learning-based models have proved to be a powerful feature extractor to perform high-accuracy prediction. However, the effectiveness of models still suffers from two issues in PWL prediction: the correlations shifting over time between PWL and other feature parameters, and the example imbalance between fluctuation examples (minority) and stable examples (majority). To address these problems, we propose a cost-sensitive mechanism to facilitate the model to learn the feature representation of later examples and fluctuation examples. By weighting the standard mean square error loss with a cost-sensitive factor, we develop a Cost-Sensitive Long Short-Term Memory (CSLSTM) model to predict the PWL of PWRs. The overall performance of the CSLSTM is assessed by a variety of evaluation metrics with the experimental data collected from a marine PWR simulator. The comparisons with the Long Short-Term Memory (LSTM) model and the Support Vector Regression (SVR) model demonstrate the effectiveness of the CSLSTM.