期刊文献+
共找到366篇文章
< 1 2 19 >
每页显示 20 50 100
Physics-informed neural network-based petroleum reservoir simulation with sparse data using domain decomposition
1
作者 Jiang-Xia Han Liang Xue +4 位作者 Yun-Sheng Wei Ya-Dong Qi Jun-Lei Wang Yue-Tian Liu Yu-Qi Zhang 《Petroleum Science》 SCIE EI CAS CSCD 2023年第6期3450-3460,共11页
Recent advances in deep learning have expanded new possibilities for fluid flow simulation in petroleum reservoirs.However,the predominant approach in existing research is to train neural networks using high-fidelity ... Recent advances in deep learning have expanded new possibilities for fluid flow simulation in petroleum reservoirs.However,the predominant approach in existing research is to train neural networks using high-fidelity numerical simulation data.This presents a significant challenge because the sole source of authentic wellbore production data for training is sparse.In response to this challenge,this work introduces a novel architecture called physics-informed neural network based on domain decomposition(PINN-DD),aiming to effectively utilize the sparse production data of wells for reservoir simulation with large-scale systems.To harness the capabilities of physics-informed neural networks(PINNs)in handling small-scale spatial-temporal domain while addressing the challenges of large-scale systems with sparse labeled data,the computational domain is divided into two distinct sub-domains:the well-containing and the well-free sub-domain.Moreover,the two sub-domains and the interface are rigorously constrained by the governing equations,data matching,and boundary conditions.The accuracy of the proposed method is evaluated on two problems,and its performance is compared against state-of-the-art PINNs through numerical analysis as a benchmark.The results demonstrate the superiority of PINN-DD in handling large-scale reservoir simulation with limited data and show its potential to outperform conventional PINNs in such scenarios. 展开更多
关键词 Physical-informed neural networks Fluid flow simulation sparse data Domain decomposition
下载PDF
Sparse Seismic Data Reconstruction Based on a Convolutional Neural Network Algorithm
2
作者 HOU Xinwei TONG Siyou +3 位作者 WANG Zhongcheng XU Xiugang PENG Yin WANG Kai 《Journal of Ocean University of China》 SCIE CAS CSCD 2023年第2期410-418,共9页
At present,the acquisition of seismic data is developing toward high-precision and high-density methods.However,complex natural environments and cultural factors in many exploration areas cause difficulties in achievi... At present,the acquisition of seismic data is developing toward high-precision and high-density methods.However,complex natural environments and cultural factors in many exploration areas cause difficulties in achieving uniform and intensive acquisition,which makes complete seismic data collection impossible.Therefore,data reconstruction is required in the processing link to ensure imaging accuracy.Deep learning,as a new field in rapid development,presents clear advantages in feature extraction and modeling.In this study,the convolutional neural network deep learning algorithm is applied to seismic data reconstruction.Based on the convolutional neural network algorithm and combined with the characteristics of seismic data acquisition,two training strategies of supervised and unsupervised learning are designed to reconstruct sparse acquisition seismic records.First,a supervised learning strategy is proposed for labeled data,wherein the complete seismic data are segmented as the input of the training set and are randomly sampled before each training,thereby increasing the number of samples and the richness of features.Second,an unsupervised learning strategy based on large samples is proposed for unlabeled data,and the rolling segmentation method is used to update(pseudo)labels and training parameters in the training process.Through the reconstruction test of simulated and actual data,the deep learning algorithm based on a convolutional neural network shows better reconstruction quality and higher accuracy than compressed sensing based on Curvelet transform. 展开更多
关键词 deep learning convolutional neural network seismic data reconstruction compressed sensing sparse collection supervised learning unsupervised learning
下载PDF
Application of graph neural network and feature information enhancement in relation inference of sparse knowledge graph
3
作者 Hai-Tao Jia Bo-Yang Zhang +4 位作者 Chao Huang Wen-Han Li Wen-Bo Xu Yu-Feng Bi Li Ren 《Journal of Electronic Science and Technology》 EI CAS CSCD 2023年第2期44-54,共11页
At present,knowledge embedding methods are widely used in the field of knowledge graph(KG)reasoning,and have been successfully applied to those with large entities and relationships.However,in research and production ... At present,knowledge embedding methods are widely used in the field of knowledge graph(KG)reasoning,and have been successfully applied to those with large entities and relationships.However,in research and production environments,there are a large number of KGs with a small number of entities and relations,which are called sparse KGs.Limited by the performance of knowledge extraction methods or some other reasons(some common-sense information does not appear in the natural corpus),the relation between entities is often incomplete.To solve this problem,a method of the graph neural network and information enhancement is proposed.The improved method increases the mean reciprocal rank(MRR)and Hit@3 by 1.6%and 1.7%,respectively,when the sparsity of the FB15K-237 dataset is 10%.When the sparsity is 50%,the evaluation indexes MRR and Hit@10 are increased by 0.8%and 1.8%,respectively. 展开更多
关键词 Feature information enhancement Graph neural network Natural language processing sparse knowledge graph(KG)inference
下载PDF
Sparse Autoencoder-based Multi-head Deep Neural Networks for Machinery Fault Diagnostics with Detection of Novelties 被引量:2
4
作者 Zhe Yang Dejan Gjorgjevikj +3 位作者 Jianyu Long Yanyang Zi Shaohui Zhang Chuan Li 《Chinese Journal of Mechanical Engineering》 SCIE EI CAS CSCD 2021年第3期146-157,共12页
Supervised fault diagnosis typically assumes that all the types of machinery failures are known.However,in practice unknown types of defect,i.e.,novelties,may occur,whose detection is a challenging task.In this paper,... Supervised fault diagnosis typically assumes that all the types of machinery failures are known.However,in practice unknown types of defect,i.e.,novelties,may occur,whose detection is a challenging task.In this paper,a novel fault diagnostic method is developed for both diagnostics and detection of novelties.To this end,a sparse autoencoder-based multi-head Deep Neural Network(DNN)is presented to jointly learn a shared encoding representation for both unsupervised reconstruction and supervised classification of the monitoring data.The detection of novelties is based on the reconstruction error.Moreover,the computational burden is reduced by directly training the multi-head DNN with rectified linear unit activation function,instead of performing the pre-training and fine-tuning phases required for classical DNNs.The addressed method is applied to a benchmark bearing case study and to experimental data acquired from a delta 3D printer.The results show that its performance is satisfactory both in detection of novelties and fault diagnosis,outperforming other state-of-the-art methods.This research proposes a novel fault diagnostics method which can not only diagnose the known type of defect,but also detect unknown types of defects. 展开更多
关键词 Deep learning Fault diagnostics Novelty detection Multi-head deep neural network sparse autoencoder
下载PDF
Alzheimer’s Disease Stage Classification Using a Deep Transfer Learning and Sparse Auto Encoder Method 被引量:1
5
作者 Deepthi K.Oommen J.Arunnehru 《Computers, Materials & Continua》 SCIE EI 2023年第7期793-811,共19页
Alzheimer’s Disease(AD)is a progressive neurological disease.Early diagnosis of this illness using conventional methods is very challenging.Deep Learning(DL)is one of the finest solutions for improving diagnostic pro... Alzheimer’s Disease(AD)is a progressive neurological disease.Early diagnosis of this illness using conventional methods is very challenging.Deep Learning(DL)is one of the finest solutions for improving diagnostic procedures’performance and forecast accuracy.The disease’s widespread distribution and elevated mortality rate demonstrate its significance in the older-onset and younger-onset age groups.In light of research investigations,it is vital to consider age as one of the key criteria when choosing the subjects.The younger subjects are more susceptible to the perishable side than the older onset.The proposed investigation concentrated on the younger onset.The research used deep learning models and neuroimages to diagnose and categorize the disease at its early stages automatically.The proposed work is executed in three steps.The 3D input images must first undergo image pre-processing using Weiner filtering and Contrast Limited Adaptive Histogram Equalization(CLAHE)methods.The Transfer Learning(TL)models extract features,which are subsequently compressed using cascaded Auto Encoders(AE).The final phase entails using a Deep Neural Network(DNN)to classify the phases of AD.The model was trained and tested to classify the five stages of AD.The ensemble ResNet-18 and sparse autoencoder with DNN model achieved an accuracy of 98.54%.The method is compared to state-of-the-art approaches to validate its efficacy and performance. 展开更多
关键词 Alzheimer’s disease mild cognitive impairment Weiner filter contrast limited adaptive histogram equalization transfer learning sparse autoencoder deep neural network
下载PDF
Learning sparse and smooth functions by deep Sigmoid nets
6
作者 LIU Xia 《Applied Mathematics(A Journal of Chinese Universities)》 SCIE CSCD 2023年第2期293-309,共17页
To pursue the outperformance of deep nets in learning,we construct a deep net with three hidden layers and prove that,implementing the empirical risk minimization(ERM)on this deep net,the estimator can theoretically r... To pursue the outperformance of deep nets in learning,we construct a deep net with three hidden layers and prove that,implementing the empirical risk minimization(ERM)on this deep net,the estimator can theoretically realize the optimal learning rates without the classical saturation problem.In other words,deepening the networks with only three hidden layers can overcome the saturation and not degrade the optimal learning rates.The obtained results underlie the success of deep nets and provide a theoretical guidance for deep learning. 展开更多
关键词 GENERALIZATION deep learning deep neural networks learning rate sparse
下载PDF
Multi-Layer Deep Sparse Representation for Biological Slice Image Inpainting
7
作者 Haitao Hu Hongmei Ma Shuli Mei 《Computers, Materials & Continua》 SCIE EI 2023年第9期3813-3832,共20页
Biological slices are an effective tool for studying the physiological structure and evolutionmechanism of biological systems.However,due to the complexity of preparation technology and the presence of many uncontroll... Biological slices are an effective tool for studying the physiological structure and evolutionmechanism of biological systems.However,due to the complexity of preparation technology and the presence of many uncontrollable factors during the preparation processing,leads to problems such as difficulty in preparing slice images and breakage of slice images.Therefore,we proposed a biological slice image small-scale corruption inpainting algorithm with interpretability based on multi-layer deep sparse representation,achieving the high-fidelity reconstruction of slice images.We further discussed the relationship between deep convolutional neural networks and sparse representation,ensuring the high-fidelity characteristic of the algorithm first.A novel deep wavelet dictionary is proposed that can better obtain image prior and possess learnable feature.And multi-layer deep sparse representation is used to implement dictionary learning,acquiring better signal expression.Compared with methods such as NLABH,Shearlet,Partial Differential Equation(PDE),K-Singular Value Decomposition(K-SVD),Convolutional Sparse Coding,and Deep Image Prior,the proposed algorithm has better subjective reconstruction and objective evaluation with small-scale image data,which realized high-fidelity inpainting,under the condition of small-scale image data.And theOn2-level time complexitymakes the proposed algorithm practical.The proposed algorithm can be effectively extended to other cross-sectional image inpainting problems,such as magnetic resonance images,and computed tomography images. 展开更多
关键词 Deep sparse representation image inpainting convolutional sparse modelling deep neural network
下载PDF
Anomaly-Based Intrusion DetectionModel Using Deep Learning for IoT Networks
8
作者 Muaadh A.Alsoufi Maheyzah Md Siraj +4 位作者 Fuad A.Ghaleb Muna Al-Razgan Mahfoudh Saeed Al-Asaly Taha Alfakih Faisal Saeed 《Computer Modeling in Engineering & Sciences》 SCIE EI 2024年第10期823-845,共23页
The rapid growth of Internet of Things(IoT)devices has brought numerous benefits to the interconnected world.However,the ubiquitous nature of IoT networks exposes them to various security threats,including anomaly int... The rapid growth of Internet of Things(IoT)devices has brought numerous benefits to the interconnected world.However,the ubiquitous nature of IoT networks exposes them to various security threats,including anomaly intrusion attacks.In addition,IoT devices generate a high volume of unstructured data.Traditional intrusion detection systems often struggle to cope with the unique characteristics of IoT networks,such as resource constraints and heterogeneous data sources.Given the unpredictable nature of network technologies and diverse intrusion methods,conventional machine-learning approaches seem to lack efficiency.Across numerous research domains,deep learning techniques have demonstrated their capability to precisely detect anomalies.This study designs and enhances a novel anomaly-based intrusion detection system(AIDS)for IoT networks.Firstly,a Sparse Autoencoder(SAE)is applied to reduce the high dimension and get a significant data representation by calculating the reconstructed error.Secondly,the Convolutional Neural Network(CNN)technique is employed to create a binary classification approach.The proposed SAE-CNN approach is validated using the Bot-IoT dataset.The proposed models exceed the performance of the existing deep learning approach in the literature with an accuracy of 99.9%,precision of 99.9%,recall of 100%,F1 of 99.9%,False Positive Rate(FPR)of 0.0003,and True Positive Rate(TPR)of 0.9992.In addition,alternative metrics,such as training and testing durations,indicated that SAE-CNN performs better. 展开更多
关键词 IOT anomaly intrusion detection deep learning sparse autoencoder convolutional neural network
下载PDF
Source Recovery in Underdetermined Blind Source Separation Based on Artificial Neural Network 被引量:3
9
作者 Weihong Fu Bin Nong +2 位作者 Xinbiao Zhou Jun Liu Changle Li 《China Communications》 SCIE CSCD 2018年第1期140-154,共15页
We propose a novel source recovery algorithm for underdetermined blind source separation, which can result in better accuracy and lower computational cost. On the basis of the model of underdetermined blind source sep... We propose a novel source recovery algorithm for underdetermined blind source separation, which can result in better accuracy and lower computational cost. On the basis of the model of underdetermined blind source separation, the artificial neural network with single-layer perceptron is introduced into the proposed algorithm. Source signals are regarded as the weight vector of single-layer perceptron, and approximate ι~0-norm is taken into account for output error decision rule of the perceptron, which leads to the sparse recovery. Then the procedure of source recovery is adjusting the weight vector of the perceptron. What's more, the optimal learning factor is calculated and a descent sequence of smoothed parameter is used during iteration, which improves the performance and significantly decreases computational complexity of the proposed algorithm. The simulation results reveal that the algorithm proposed can recover the source signal with high precision, while it requires lower computational cost. 展开更多
关键词 underdetermined blind source separation ι~0-norm artificial neural network sparse reconstruction
下载PDF
Performance Evaluation of Super-Resolution Methods Using Deep-Learning and Sparse-Coding for Improving the Image Quality of Magnified Images in Chest Radiographs
10
作者 Kensuke Umehara Junko Ota +4 位作者 Naoki Ishimaru Shunsuke Ohno Kentaro Okamoto Takanori Suzuki Takayuki Ishida 《Open Journal of Medical Imaging》 2017年第3期100-111,共12页
Purpose: To detect small diagnostic signals such as lung nodules in chest radiographs, radiologists magnify a region-of-interest using linear interpolation methods. However, such methods tend to generate over-smoothed... Purpose: To detect small diagnostic signals such as lung nodules in chest radiographs, radiologists magnify a region-of-interest using linear interpolation methods. However, such methods tend to generate over-smoothed images with artifacts that can make interpretation difficult. The purpose of this study was to investigate the effectiveness of super-resolution methods for improving the image quality of magnified chest radiographs. Materials and Methods: A total of 247 chest X-rays were sampled from the JSRT database, then divided into 93 training cases with non-nodules and 154 test cases with lung nodules. We first trained two types of super-resolution methods, sparse-coding super-resolution (ScSR) and super-resolution convolutional neural network (SRCNN). With the trained super-resolution methods, the high-resolution image was then reconstructed using the super-resolution methods from a low-resolution image that was down-sampled from the original test image. We compared the image quality of the super-resolution methods and the linear interpolations (nearest neighbor and bilinear interpolations). For quantitative evaluation, we measured two image quality metrics: peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). For comparative evaluation of the super-resolution methods, we measured the computation time per image. Results: The PSNRs and SSIMs for the ScSR and the SRCNN schemes were significantly higher than those of the linear interpolation methods (p p p Conclusion: Super-resolution methods provide significantly better image quality than linear interpolation methods for magnified chest radiograph images. Of the two tested schemes, the SRCNN scheme processed the images fastest;thus, SRCNN could be clinically superior for processing radiographs in terms of both image quality and processing speed. 展开更多
关键词 Deep LEARNING SUPER-RESOLUTION SUPER-RESOLUTION Convolutional neural network (SRCNN) sparse-Coding SUPER-RESOLUTION (ScSR) CHEST X-Ray
下载PDF
融合稀疏八叉树与卷积神经网络的汽车风阻系数预测 被引量:1
11
作者 王刚 张瑞昊 +2 位作者 刘学龙 袁海东 韩旭 《计算力学学报》 CAS CSCD 北大核心 2024年第1期58-65,共8页
针对汽车风阻系数预测研究中参数化方法难以准确表征汽车外造型的难题,提出融合稀疏八叉树与卷积神经网络的汽车风阻系数预测方法。将汽车外造型按照八叉树结构离散,使用平均法向量对离散的复杂曲面进行简化,利用卷积神经网络对八叉树... 针对汽车风阻系数预测研究中参数化方法难以准确表征汽车外造型的难题,提出融合稀疏八叉树与卷积神经网络的汽车风阻系数预测方法。将汽车外造型按照八叉树结构离散,使用平均法向量对离散的复杂曲面进行简化,利用卷积神经网络对八叉树形式的汽车外造型进行特征提取,进而对汽车风阻系数进行快速预测。通过改变卷积层数与全连接层数,研究了不同卷积神经网络结构对风阻系数预测精度的影响。与参数化方法相比,本文提出的外造型表示方法能更好地描述模型细节,构建的卷积神经网络结构对风阻系数预测的最小相对误差为1.453%,且计算速度是CFD仿真的1620倍,具有较高的精度及计算效率。 展开更多
关键词 风阻系数 深度学习 稀疏八叉树 卷积神经网络 汽车
下载PDF
多任务联合学习的图卷积神经网络推荐 被引量:1
12
作者 王永贵 邹赫宇 《计算机工程与应用》 CSCD 北大核心 2024年第4期306-314,共9页
基于图神经网络的协同过滤推荐可以更有效地挖掘用户项目之间的交互信息,但其性能依然受到数据稀疏和表征学习质量不高问题的影响,因此提出一种多任务联合学习的图卷积神经网络推荐(multi-task joint learning for graph convolutional ... 基于图神经网络的协同过滤推荐可以更有效地挖掘用户项目之间的交互信息,但其性能依然受到数据稀疏和表征学习质量不高问题的影响,因此提出一种多任务联合学习的图卷积神经网络推荐(multi-task joint learning for graph convolutional neural network recommendations,MTJL-GCN)模型。利用图神经网络在用户-项目交互图上所聚集到的同质结构信息与初始嵌入信息形成结构邻居关系,设计节点邻居关系的对比学习辅助任务来缓解数据稀疏问题;向节点的原始表征添加随机的统一噪声进行表征级数据增强,构建节点表征关系的对比学习辅助任务,并提出直接优化对齐性和均匀性两个属性的学习目标来提高表征学习质量;将图协同过滤推荐任务与对比学习辅助任务和直接优化学习目标进行联合训练,从而提升推荐性能。在Amazon-books和Yelp2018两个公开数据集上进行实验,该模型在Recall@k和NDCG@k两个推荐性能指标上的表现均优于基线模型,证明了MTJL-GCN模型的有效性。 展开更多
关键词 推荐算法 图卷积神经网络 对比学习 表征学习 数据稀疏 协同过滤
下载PDF
AR-MED共振特征增强的风电齿轮箱故障诊断
13
作者 孙抗 史晓玉 +1 位作者 赵来军 杨明 《组合机床与自动化加工技术》 北大核心 2024年第8期163-167,174,共6页
针对风电齿轮箱故障时脉冲成分往往淹没在其他频率分量中,早期故障特征难以有效提取的问题,提出一种自回归最小熵解卷积(AR-MED)共振特征增强的风电齿轮箱故障诊断方法,并结合一维卷积神经网络(1DCNN),实现齿轮箱高精度故障诊断。首先,... 针对风电齿轮箱故障时脉冲成分往往淹没在其他频率分量中,早期故障特征难以有效提取的问题,提出一种自回归最小熵解卷积(AR-MED)共振特征增强的风电齿轮箱故障诊断方法,并结合一维卷积神经网络(1DCNN),实现齿轮箱高精度故障诊断。首先,使用共振稀疏分解算法(RSSD)将振动信号分解成含有噪声和谐波成分的高共振分量和含有故障冲击成分的低共振分量;其次,对低共振分量使用自回归最小熵解卷积运算,增强低共振分量中微弱的周期性冲击成分;最后,构建自回归最小熵解卷积共振特征增强的1DCNN模型,将分解得到的谐波分量和周期性冲击分量进行特征融合以及有针对的训练和分类。实验结果表明,与现有故障诊断模型相比,所提方法在提取风电齿轮箱的故障特征信息以及提高故障诊断精度方面具有有效性和优越性。 展开更多
关键词 共振稀疏分解 自回归最小熵解卷积 特征增强 一维卷积神经网络 风电齿轮箱
下载PDF
基于多特征优化的PolSAR数据农作物精细分类方法
14
作者 郭交 王鹤颖 +2 位作者 项诗雨 连嘉茜 王辉 《农业机械学报》 EI CAS CSCD 北大核心 2024年第9期275-285,共11页
农作物精细分类在农业资源调查、农作物种植结构监管等诸多领域具有重要意义。极化合成孔径雷达(Polarimetric synthetic aperture radar,PolSAR)能够有效探测伪装和穿透掩盖物,提取多种散射特征信息,获取覆盖农作物生长关键物候阶段的... 农作物精细分类在农业资源调查、农作物种植结构监管等诸多领域具有重要意义。极化合成孔径雷达(Polarimetric synthetic aperture radar,PolSAR)能够有效探测伪装和穿透掩盖物,提取多种散射特征信息,获取覆盖农作物生长关键物候阶段的连续时序信息,有效提升表达作物遥感特征的丰富度,在农作物分类中独具优势。但多时相和多特征的引入必然导致模型运算量剧增,不利于工程应用。针对上述问题,本文提出了一种基于多特征优化的PolSAR数据农作物精细分类方法,首先对PolSAR数据进行多种极化目标分解及参数提取以获得多个散射特征;然后使用基于栈式稀疏自编码网络和ReliefF优选的方法进行特征增强与优化,获取最优特征集;最后构建具有2个分支结构的卷积神经网络,融合不同卷积深度输出的特征,完成农作物的高精度分类。通过对单时相数据的特征分析、单时相数据初步分类实验和多时相数据不同特征集结合分类器的对比实验,证明本文所提方法能够在低维特征输入的前提下,最大程度提取不同作物之间的差异性特征,准确高效地实现对农作物的精细分类,最高分类精度和Kappa系数分别达到97.69%和97.24%。 展开更多
关键词 农作物分类 POLSAR 栈式稀疏自编码网络 RELIEFF 卷积神经网络
下载PDF
基于卷积特征提取及深度降噪网络的大规模MIMO系统信号检测
15
作者 申滨 涂媛媛 +1 位作者 阳建 金龙康 《信号处理》 CSCD 北大核心 2024年第6期1030-1040,共11页
传统多输入多输出(Multiple-Input Multiple-Output,MIMO)信号检测算法受到天线数量和收发天线比例的限制,一般仅适用于少量天线、收发天线比例较低的情况。本文提出一种基于深度学习(Deep Learning,DL)的稀疏连接卷积降噪网络模型,用... 传统多输入多输出(Multiple-Input Multiple-Output,MIMO)信号检测算法受到天线数量和收发天线比例的限制,一般仅适用于少量天线、收发天线比例较低的情况。本文提出一种基于深度学习(Deep Learning,DL)的稀疏连接卷积降噪网络模型,用于大规模MIMO系统上行链路信号检测。首先,通过简化经典的检测网络(Detection Network,DetNet),改进ScNet(Sparsely Connected Neural Network)检测算法,引入卷积神经网络(Convolutional NeuralNetworks,CNN)对三通道输入数据提取特征以减少训练参数,提出一种SConv(SparselyConnected Convolutional Neural Network)检测算法。与DetNet算法相比,该算法可同时降低计算复杂度和提高检测精度。在此基础上,进一步基于CNN构建信号降噪模块,并嵌入SConv网络,提出一种卷积神经降噪(Sparsely Connected Convolutional Denoising,SConv-D)网络辅助的大规模MIMO检测算法。此算法检测过程分为两级,第一级由SConv算法提供初始估计值,再将初始估计值作为降噪过程的输入,并由此构成算法第二级。实验结果表明,本文提出的SConv-D算法适用于QPSK、4QAM及16QAM等多种信号调制模式,在高阶调制模式下获得的性能增益尤为明显。此外,该算法能够适应各种比例的收发天线及数量规模的系统配置,尤其是在收发天线数量相等的情况下亦能获得更优的性能。本文算法还克服了MMNet在高阶调制情况下的性能平台效应,在16QAM调制、收发天线数量相等的情况下,SConv-D在10^(-2)误比特率上能获得接近2 dB的性能增益。 展开更多
关键词 大规模MIMO 深度学习 稀疏连接 卷积神经网络 降噪
下载PDF
NM-SpMM:面向国产异构向量处理器的半结构化稀疏矩阵乘算法
16
作者 姜晶菲 何源宏 +2 位作者 许金伟 许诗瑶 钱希福 《计算机工程与科学》 CSCD 北大核心 2024年第7期1141-1150,共10页
深度神经网络在自然语言处理、计算机视觉等领域取得了优异的成果,由于智能应用处理数据规模的增长和大模型的快速发展,对深度神经网络的推理性能要求越来越高,N∶M半结构化稀疏化技术成为平衡算力需求和应用效果的热点技术之一。国产... 深度神经网络在自然语言处理、计算机视觉等领域取得了优异的成果,由于智能应用处理数据规模的增长和大模型的快速发展,对深度神经网络的推理性能要求越来越高,N∶M半结构化稀疏化技术成为平衡算力需求和应用效果的热点技术之一。国产异构向量处理器FT-M7032为智能模型处理中的数据并行和指令并行开发提供了较大空间。针对N∶M半结构化稀疏模型计算稀疏模式多样性,提出了一种面向FT-M7032的可灵活配置的稀疏矩阵乘算法NM-SpMM。NM-SpMM设计了一种高效的压缩偏移地址稀疏编码格式COA,避免了半结构化参数配置对稀疏数据访存计算的影响。基于COA编码,NM-SpMM对不同维度稀疏矩阵计算进行了细粒度优化。在FT-M7032单核上的实验结果表明,相较于稠密矩阵乘,NM-SpMM能获得1.73~21.00倍的加速,相较于采用CuSPARSE稀疏计算库的NVIDIA V100 GPU,能获得0.04~1.04倍的加速。 展开更多
关键词 深度神经网络 图形处理器 向量处理器 稀疏矩阵乘 流水线
下载PDF
基于黑客画像的网络攻击者识别方法
17
作者 徐雅斌 王振超 庄唯 《计算机工程与设计》 北大核心 2024年第6期1624-1630,共7页
为能够准确、快速识别网络攻击者,提出一种基于黑客画像的网络攻击者识别方法。构建将稀疏自编码器和贝叶斯神经网络相结合的SAE-BNN模型,检测不同攻击类型的恶意流量;针对不同的恶意流量,通过提取黑客属性特征、流量特征、时间特征和... 为能够准确、快速识别网络攻击者,提出一种基于黑客画像的网络攻击者识别方法。构建将稀疏自编码器和贝叶斯神经网络相结合的SAE-BNN模型,检测不同攻击类型的恶意流量;针对不同的恶意流量,通过提取黑客属性特征、流量特征、时间特征和相似性特征,与事先建立的黑客画像库中的黑客画像进行匹配。如果与某个黑客画像完全匹配,则由此确定该黑客的身份。当不能与黑客画像库中的任何黑客画像进行匹配时,将该黑客的特征作为标签,构建新的黑客画像,并更新画像库。实验结果表明,提出的异常流量识别方法在精度、召回率、F1值和准确率上均有提升。基于黑客画像的黑客识别算法与常规方法相比,极大提高了识别效率。 展开更多
关键词 稀疏自编码器 贝叶斯神经网络 网络黑客 黑客画像 黑客特征 黑客匹配 恶意流量
下载PDF
融合选择性稀疏采样的细粒度图像分类
18
作者 孙红 陈玉娟 宋冬豪 《小型微型计算机系统》 CSCD 北大核心 2024年第6期1460-1465,共6页
常用的细粒度分类方法通过提取局部信息学习细粒度特征,容易忽视周围环境因素影响问题,造成分类精度下降.针对这一问题提出了一个简单有效的框架,称为选择性稀疏采样.通过类峰值响应产生稀疏注意定位有信息的对象部分,根据图像内容选择... 常用的细粒度分类方法通过提取局部信息学习细粒度特征,容易忽视周围环境因素影响问题,造成分类精度下降.针对这一问题提出了一个简单有效的框架,称为选择性稀疏采样.通过类峰值响应产生稀疏注意定位有信息的对象部分,根据图像内容选择动态数量的稀疏注意,生成判别性和补充性两个分支进行视觉表示,使得特征部分和全局信息相辅相成.对于容易产生混淆的部分,引入了一个“梯度增强”损失,只关注每个样本的混淆类,为补充性分支提供更多的细节特征.通过实验结果表明,该方法在常用数据集的基准测试中分别达到了88.6%,92.8%和94.8%的精确度,验证了该方法的有效性. 展开更多
关键词 细粒度图像分类 选择稀疏采样 类峰值响应 梯度增强 卷积神经网络
下载PDF
时空图卷积网络的骨架识别硬件加速器设计
19
作者 谭会生 严舒琪 杨威 《电子测量技术》 北大核心 2024年第11期36-43,共8页
随着人工智能技术的不断发展,神经网络的数据规模逐渐扩大,神经网络的计算量也迅速攀升。为了减少时空图卷积神经网络的计算量,降低硬件实现的资源消耗,提升人体骨架识别时空图卷积神经网络(ST-GCN)实际应用系统的处理速度,利用现场可... 随着人工智能技术的不断发展,神经网络的数据规模逐渐扩大,神经网络的计算量也迅速攀升。为了减少时空图卷积神经网络的计算量,降低硬件实现的资源消耗,提升人体骨架识别时空图卷积神经网络(ST-GCN)实际应用系统的处理速度,利用现场可编程门阵列(FPGA),设计开发了一个基于时空图卷积神经网络的骨架识别硬件加速器。通过对原网络模型进行结构优化与数据量化,减少了FPGA实现约75%的计算量;利用邻接矩阵稀疏性的特点,提出了一种稀疏性矩阵乘加运算的优化方法,减少了约60%的乘法器资源消耗。经过对人体骨架识别实验验证,结果表明,在时钟频率100 MHz下,相较于CPU,FPGA加速ST-GCN单元,加速比达到30.53;FPGA加速人体骨架识别,加速比达到6.86。 展开更多
关键词 人体骨架识别 时空图卷积神经网络(ST-GCN) 硬件加速器 现场可编程门阵列(FPGA) 稀疏矩阵乘加运算硬件优化
原文传递
面向稀疏卷积神经网络的CGRA加速器研究 被引量:1
20
作者 谭龙 严明玉 +3 位作者 吴欣欣 李文明 吴海彬 范东睿 《高技术通讯》 CAS 北大核心 2024年第2期173-186,共14页
本文针对规模日益增长和演变迅速的稀疏卷积神经网络(CNN)应用,提出一款高能效且灵活的加速结构DyCNN来提升其性能和能效。DyCNN基于兼具灵活性和高能效的粗粒度可重构架构(CGRA)设计,可以利用其指令的高并行性来高效支持CNN的操作。Dy... 本文针对规模日益增长和演变迅速的稀疏卷积神经网络(CNN)应用,提出一款高能效且灵活的加速结构DyCNN来提升其性能和能效。DyCNN基于兼具灵活性和高能效的粗粒度可重构架构(CGRA)设计,可以利用其指令的高并行性来高效支持CNN的操作。DyCNN使用基于数据感知的指令动态过滤机制来滤除各计算单元中由于稀疏CNN中权值静态稀疏性和激活值动态稀疏性产生的大量无效计算和访存指令,使它们能像执行稠密网络一样高效复用一组指令。此外DyCNN利用基于负载感知的动静结合负载调度策略解决了稀疏导致的负载不均衡问题。实验结果表明,DyCNN运行稀疏CNN与运行密集CNN相比实现了平均1.69倍性能提升和3.04倍能效提升,比先进的GPU(cuSPARSE)和Cambricon-X上的解决方案分别实现了2.78倍、1.48倍性能提升和35.62倍、1.17倍能效提升。 展开更多
关键词 稀疏卷积神经网络(CNN) 专用加速结构 粗粒度可重构架构(CGRA) 动态指令过滤 动态负载调度
下载PDF
上一页 1 2 19 下一页 到第
使用帮助 返回顶部