亚洲男人的天堂2018av,欧美草比,久久久久久免费视频精选,国色天香在线看免费,久久久久亚洲av成人片仓井空

This work extends the paradigm of evolutional deep neural networks (EDNNs) to solving parametric time-dependent partial differential equations (PDEs) on domains with geometric structure. By introducing positional embeddings based on eigenfunctions of the Laplace-Beltrami operator, geometric properties are encoded intrinsically and Dirichlet, Neumann and periodic boundary conditions of the PDE solution are enforced directly through the neural network architecture. The proposed embeddings lead to improved error convergence for static PDEs and extend EDNNs towards computational domains of realistic complexity. Several steps are taken to improve performance of EDNNs: Solving the EDNN update equation with a Krylov solver avoids the explicit assembly of Jacobians and enables scaling to larger neural networks. Computational efficiency is further improved by an ad-hoc active sampling scheme that uses the PDE dynamics to effectively sample collocation points. A modified linearly implicit Rosenbrock method is proposed to alleviate the time step requirements of stiff PDEs. Lastly, a completely training-free approach, which automatically enforces initial conditions and only requires time integration, is compared against EDNNs that are trained on the initial conditions. We report results for the Korteweg-de Vries equation, a nonlinear heat equation and (nonlinear) advection-diffusion problems on domains with and without holes and various boundary conditions, to demonstrate the effectiveness of the method. The numerical results highlight EDNNs as a promising surrogate model for parametrized PDEs with slow decaying Kolmogorov n-width.

The emergence of large-scale wireless networks with partially-observable and time-varying dynamics has imposed new challenges on the design of optimal control policies. This paper studies efficient scheduling algorithms for wireless networks subject to generalized interference constraint, where mean arrival and mean service rates are unknown and non-stationary. This model exemplifies realistic edge devices' characteristics of wireless communication in modern networks. We propose a novel algorithm termed MW-UCB for generalized wireless network scheduling, which is based on the Max-Weight policy and leverages the Sliding-Window Upper-Confidence Bound to learn the channels' statistics under non-stationarity. MW-UCB is provably throughput-optimal under mild assumptions on the variability of mean service rates. Specifically, as long as the total variation in mean service rates over any time period grows sub-linearly in time, we show that MW-UCB can achieve the stability region arbitrarily close to the stability region of the class of policies with full knowledge of the channel statistics. Extensive simulations validate our theoretical results and demonstrate the favorable performance of MW-UCB.

With the rapid proliferation of smart mobile devices, federated learning (FL) has been widely considered for application in wireless networks for distributed model training. However, data heterogeneity, e.g., non-independently identically distributions and different sizes of training data among clients, poses major challenges to wireless FL. Limited communication resources complicate the implementation of fair scheduling which is required for training on heterogeneous data, and further deteriorate the overall performance. To address this issue, this paper focuses on performance analysis and optimization for wireless FL, considering data heterogeneity, combined with wireless resource allocation. Specifically, we first develop a closed-form expression for an upper bound on the FL loss function, with a particular emphasis on data heterogeneity described by a dataset size vector and a data divergence vector. Then we formulate the loss function minimization problem, under constraints on long-term energy consumption and latency, and jointly optimize client scheduling, resource allocation, and the number of local training epochs (CRE). Next, via the Lyapunov drift technique, we transform the CRE optimization problem into a series of tractable problems. Extensive experiments on real-world datasets demonstrate that the proposed algorithm outperforms other benchmarks in terms of the learning accuracy and energy consumption.

With the rise of the popularity and usage of neural networks, trustworthy uncertainty estimation is becoming increasingly essential. One of the most prominent uncertainty estimation methods is Deep Ensembles (Lakshminarayanan et al., 2017) . A classical parametric model has uncertainty in the parameters due to the fact that the data on which the model is build is a random sample. A modern neural network has an additional uncertainty component since the optimization of the network is random. Lakshminarayanan et al. (2017) noted that Deep Ensembles do not incorporate the classical uncertainty induced by the effect of finite data. In this paper, we present a computationally cheap extension of Deep Ensembles for the regression setting, called Bootstrapped Deep Ensembles, that explicitly takes this classical effect of finite data into account using a modified version of the parametric bootstrap. We demonstrate through an experimental study that our method significantly improves upon standard Deep Ensembles

For safety-related applications, it is crucial to produce trustworthy deep neural networks whose prediction is associated with confidence that can represent the likelihood of correctness for subsequent decision-making. Existing dense binary classification models are prone to being over-confident. To improve model calibration, we propose Adaptive Stochastic Label Perturbation (ASLP) which learns a unique label perturbation level for each training image. ASLP employs our proposed Self-Calibrating Binary Cross Entropy (SC-BCE) loss, which unifies label perturbation processes including stochastic approaches (like DisturbLabel), and label smoothing, to correct calibration while maintaining classification rates. ASLP follows Maximum Entropy Inference of classic statistical mechanics to maximise prediction entropy with respect to missing information. It performs this while: (1) preserving classification accuracy on known data as a conservative solution, or (2) specifically improves model calibration degree by minimising the gap between the prediction accuracy and expected confidence of the target training label. Extensive results demonstrate that ASLP can significantly improve calibration degrees of dense binary classification models on both in-distribution and out-of-distribution data. The code is available on //github.com/Carlisle-Liu/ASLP.

Unmanned aerial vehicle (UAV) network is a promising technology for assisting Internet-of-Things (IoT), where a UAV can use its limited service coverage to harvest and disseminate data from IoT devices with low transmission abilities. The existing UAV-assisted data harvesting and dissemination schemes largely require UAVs to frequently fly between the IoTs and access points, resulting in extra energy and time costs. To reduce both energy and time costs, a key way is to enhance the transmission performance of IoT and UAVs. In this work, we introduce collaborative beamforming into IoTs and UAVs simultaneously to achieve energy and time-efficient data harvesting and dissemination from multiple IoT clusters to remote base stations (BSs). Except for reducing these costs, another non-ignorable threat lies in the existence of the potential eavesdroppers, whereas the handling of eavesdroppers often increases the energy and time costs, resulting in a conflict with the minimization of the costs. Moreover, the importance of these goals may vary relatively in different applications. Thus, we formulate a multi-objective optimization problem (MOP) to simultaneously minimize the mission completion time, signal strength towards the eavesdropper, and total energy cost of the UAVs. We prove that the formulated MOP is an NP-hard, mixed-variable optimization, and large-scale optimization problem. Thus, we propose a swarm intelligence-based algorithm to find a set of candidate solutions with different trade-offs which can meet various requirements in a low computational complexity. We also show that swarm intelligence methods need to enhance solution initialization, solution update, and algorithm parameter update phases when dealing with mixed-variable optimization and large-scale problems. Simulation results demonstrate the proposed algorithm outperforms state-of-the-art swarm intelligence algorithms.

Pre-trained deep neural network language models such as ELMo, GPT, BERT and XLNet have recently achieved state-of-the-art performance on a variety of language understanding tasks. However, their size makes them impractical for a number of scenarios, especially on mobile and edge devices. In particular, the input word embedding matrix accounts for a significant proportion of the model's memory footprint, due to the large input vocabulary and embedding dimensions. Knowledge distillation techniques have had success at compressing large neural network models, but they are ineffective at yielding student models with vocabularies different from the original teacher models. We introduce a novel knowledge distillation technique for training a student model with a significantly smaller vocabulary as well as lower embedding and hidden state dimensions. Specifically, we employ a dual-training mechanism that trains the teacher and student models simultaneously to obtain optimal word embeddings for the student vocabulary. We combine this approach with learning shared projection matrices that transfer layer-wise knowledge from the teacher model to the student model. Our method is able to compress the BERT_BASE model by more than 60x, with only a minor drop in downstream task metrics, resulting in a language model with a footprint of under 7MB. Experimental results also demonstrate higher compression efficiency and accuracy when compared with other state-of-the-art compression techniques.

Deep convolutional neural networks (CNNs) have recently achieved great success in many visual recognition tasks. However, existing deep neural network models are computationally expensive and memory intensive, hindering their deployment in devices with low memory resources or in applications with strict latency requirements. Therefore, a natural thought is to perform model compression and acceleration in deep networks without significantly decreasing the model performance. During the past few years, tremendous progress has been made in this area. In this paper, we survey the recent advanced techniques for compacting and accelerating CNNs model developed. These techniques are roughly categorized into four schemes: parameter pruning and sharing, low-rank factorization, transferred/compact convolutional filters, and knowledge distillation. Methods of parameter pruning and sharing will be described at the beginning, after that the other techniques will be introduced. For each scheme, we provide insightful analysis regarding the performance, related applications, advantages, and drawbacks etc. Then we will go through a few very recent additional successful methods, for example, dynamic capacity networks and stochastic depths networks. After that, we survey the evaluation matrix, the main datasets used for evaluating the model performance and recent benchmarking efforts. Finally, we conclude this paper, discuss remaining challenges and possible directions on this topic.

Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs---a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DiffPool learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DiffPool yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark data sets.

Deep neural networks (DNNs) have been found to be vulnerable to adversarial examples resulting from adding small-magnitude perturbations to inputs. Such adversarial examples can mislead DNNs to produce adversary-selected results. Different attack strategies have been proposed to generate adversarial examples, but how to produce them with high perceptual quality and more efficiently requires more research efforts. In this paper, we propose AdvGAN to generate adversarial examples with generative adversarial networks (GANs), which can learn and approximate the distribution of original instances. For AdvGAN, once the generator is trained, it can generate adversarial perturbations efficiently for any instance, so as to potentially accelerate adversarial training as defenses. We apply AdvGAN in both semi-whitebox and black-box attack settings. In semi-whitebox attacks, there is no need to access the original target model after the generator is trained, in contrast to traditional white-box attacks. In black-box attacks, we dynamically train a distilled model for the black-box model and optimize the generator accordingly. Adversarial examples generated by AdvGAN on different target models have high attack success rate under state-of-the-art defenses compared to other attacks. Our attack has placed the first with 92.76% accuracy on a public MNIST black-box attack challenge.