Electroencephalography (EEG) offers non-invasive, real-time mental workload assessment, which is crucial in high-stakes domains like aviation and medicine and for advancing brain-computer interface (BCI) technologies....
详细信息
Given a set P of n points in R^d for d >= 2 and an integer parameter 0 = 3. These algorithms also work properly for the weighted variant of the problems in the same time bound. By our second approach, we present fa...
详细信息
Reconfigurable intelligent surface (RIS) is one of the promising technology for the next-generation wireless networks. The RIS reflects the received signal with phase shift introduced by reflecting elements without an...
Reconfigurable intelligent surface (RIS) is one of the promising technology for the next-generation wireless networks. The RIS reflects the received signal with phase shift introduced by reflecting elements without analog to digital conversion. In relay transmission scenario, RIS could be exploited to enhance the network coverage and improve transmission performance. In this paper, we consider RIS-assisted full-duplex (FD) relay transmission scenario. In the considered scenario, RIS and FD relay node performs relay transmission. Although the beamformed signal could be generated by phase shift of RIS, the interference from RIS to FD relay node may degrade the throughput performance. The proposed method cancels both of interference from RIS and self-interference at the FD relay node by estimating interference channels in the preparation phase before data transmission.
The landscape of deep learning has vastly expanded the frontiers of source code analysis, particularly through the utilization of structural representations such as Abstract Syntax Trees (ASTs). While these methodolog...
详细信息
As real environments are quite complex, multiple-input-multiple-output (MIMO) channels should be characterized by the multidimensional-multirelational models. Tensor linear algebra has been emerging as a powerful and ...
详细信息
The softmax function is ubiquitous in machine learning and optimization applications. Computing the full softmax evaluation of a matrix-vector product can be computationally expensive in high-dimensional settings. In ...
In the realm of deep learning, image restoration has seen considerable advancements, yet shadow removal remains a persistent challenge due to the variable size and color of shadows influenced by lighting conditions. T...
详细信息
It has become cognitive inertia to employ cross-entropy loss function in classification related tasks. In the untargeted attacks on graph structure, the gradients derived from the attack objective are the attacker'...
ISBN:
(纸本)9781713871088
It has become cognitive inertia to employ cross-entropy loss function in classification related tasks. In the untargeted attacks on graph structure, the gradients derived from the attack objective are the attacker's basis for evaluating a perturbation scheme. Previous methods use negative cross-entropy loss as the attack objective in attacking node-level classification models. However, the suitability of the cross-entropy function for constructing the untargeted attack objective has yet been discussed in previous works. This paper argues about the previous unreasonable attack objective from the perspective of budget allocation. We demonstrate theoretically and empirically that negative cross-entropy tends to produce more significant gradients from nodes with lower confidence in the labeled classes, even if the predicted classes of these nodes have been misled. To free up these inefficient attack budgets, we propose a simple attack model for untargeted attacks on graph structure based on a novel attack objective which generates unweighted gradients on graph structures that are not affected by the node confidence. By conducting experiments in gray-box poisoning attack scenarios, we demonstrate that a reasonable budget allocation can significantly improve the effectiveness of gradient-based edge perturbations without any extra hyper-parameter.
暂无评论