检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Ltaief, Hatem Alomairy, Rabab Cao, Qinglei Ren, Jie Slim, Lotfi Kurth, Thorsten Dorschner, Benedikt Bougouffa, Salim Abdelkhalak, Rached Keyes, David E. Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Saudi Arabia Computer Science & Artificial Intelligence Laboratory Massachusetts Institute of Technology United States Department of Computer Science Saint Louis University United States NVIDIA United States

We exploit the widening margin in tensor-core performance between [FP64/FP32/FP16/INT8,FP64/FP32/FP16/FP8/INT8] on NVIDIA [Ampere,Hopper] GPUs to boost the performance of output accuracy-preserving mixed-precision computation of Genome-Wide Association Studies (GWAS) of 305K patients from the UK BioBank, the largest-ever GWAS cohort studied for genetic epistasis using a multivariate approach. Tile-centric adaptive-precision linear algebraic techniques motivated by reducing data motion gain enhanced significance with low-precision GPU arithmetic. At the core of Kernel Ridge Regression (KRR) techniques for GWAS lie compute-bound cubic-complexity matrix operations that inhibit scaling to aspirational dimensions of the population, genotypes, and phenotypes. We accelerate KRR matrix generation by redesigning the computation for Euclidean distances to engage INT8 tensor cores while exploiting symmetry. We accelerate solution of the regularized KRR systems by deploying a new four-precision Cholesky-based solver, which, at 1.805 mixed-precision ExaOp/s on a nearly full Alps system, outperforms the state-of-the-art CPU-only REGENIE GWAS software by five orders of magnitude. Copyright © 2024, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

Bayesian Parameter Inference for Partially Observed SDEs driven by Fractional Brownian Motion

arXiv

引用

arXiv 2022年

作者： Maama, Mohamed Jasra, Ajay Ombao, Hernando Applied Mathematics and Computational Science Program Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia Statistics Program Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia

In this paper we consider Bayesian parameter inference for partially observed fractional Brownian motion (fBM) models. The approach we follow is to time-discretize the hidden process and then to design Markov chain Monte Carlo (MCMC) algorithms to sample from the posterior density on the parameters given data. We rely on a novel representation of the time discretization, which seeks to sample from an approximation of the posterior and then corrects via importance sampling;the approximation reduces the time (in terms of total observation time T) by O(T). This method is extended by using a multilevel MCMC method which can reduce the computational cost to achieve a given mean square error (MSE) versus using a single time discretization. Our methods are illustrated on simulated and real data. Copyright © 2022, The Authors. All rights reserved.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Synergistic Swarm Optimization Algorithm

引用

computer Modeling in engineering & sciences 2024年第6期139卷 2557-2604页

作者： Sharaf Alzoubi Laith Abualigah Mohamed Sharaf Mohammad Sh.Daoud Nima Khodadadi Heming Jia Faculty of Computer Sciences and Informatics Amman Arab UniversityAmman11953Jordan Computer Science Department Al Al-Bayt UniversityMafraq25113Jordan Department of Electrical and Computer Engineering Lebanese American UniversityByblos13-5053Lebanon Hourani Center for Applied Scientific Research Al-Ahliyya Amman UniversityAmman19328Jordan MEU Research Unit Middle East UniversityAmman11831Jordan Applied Science Research Center Applied Science Private UniversityAmman11931Jordan School of Computer Sciences Universiti Sains MalaysiaPulau Pinang11800Malaysia School of Engineering and Technology Sunway University MalaysiaPetaling Jaya27500Malaysia Department of Industrial Engineering College of EngineeringKing Saud UniversityP.O.Box 800Riyadh11421Saudi Arabia College of Engineering Al Ain UniversityAbu Dhabi112612United Arab Emirates Department of Civil and Architectural Engineering University of MiamiCoral Gables1251USA School of Information Engineering Sanming UniversitySanming365004China

This research paper presents a novel optimization method called the Synergistic Swarm Optimization Algorithm(SSOA).The SSOA combines the principles of swarmintelligence and synergistic cooperation to search for optimal solutions efficiently.A synergistic cooperation mechanism is employed,where particles exchange information and learn from each other to improve their search *** cooperation enhances the exploitation of promising regions in the search space while maintaining exploration ***,adaptive mechanisms,such as dynamic parameter adjustment and diversification strategies,are incorporated to balance exploration and *** leveraging the collaborative nature of swarm intelligence and integrating synergistic cooperation,the SSOAmethod aims to achieve superior convergence speed and solution quality performance compared to other optimization *** effectiveness of the proposed SSOA is investigated in solving the 23 benchmark functions and various engineering design *** experimental results highlight the effectiveness and potential of the SSOA method in addressing challenging optimization problems,making it a promising tool for a wide range of applications in engineering and *** codes of SSOA are available at:https://***/matlabcentral/fileexchange/153466-synergistic-swarm-optimization-algorithm.

关键词： Synergistic swarm optimization algorithm optimization algorithm metaheuristic engineering problems benchmark functions

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of LEO Satellite-assisted Caching: A Stochastic Geometry Approach

Performance Evaluation of LEO Satellite-assisted Caching: A ...

引用

International Conference on Communication technology (ICCT)

作者： Chunyi Ma Jiajie Xu Jianhua Yang Mustafa A. Kishk School of Automation Northwestern Polytechnical University Xi’an China Computer Electrical and Mathematical Science and Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia Department of Electronic Engineering Maynooth University Maynooth Ireland

ISBN: (数字)9798350363760

ISBN: (纸本)9798350363777

In order to satisfy the rapidly growing demand for data traffic from massive users, this study investigates a space-ground LEO satellite-assisted edge caching network framework and analyses the downlink transmission performance of this edge caching network. The aim is to enhance the quality of service provided by the edge caching network to users through the utilisation of LEO satellites. We use stochastic geometry to analyse the downlink performance. The spatial locations of the LEO satellites are modeled as Binomial Point Process (BPP) distribution deployed at a specific space altitude, and the locations of the terrestrial cloud servers (CSs) and mobile users (MUs) are modelled as independent Poisson Point Process (PPP), respectively. We establish the association strategy of typical MUs to LEO satellites (SATs) and CSs, based on which we derive the coverage probability analytical expression and average delay numerical solution for this network system. Finally, we verify the correctness of the numerical result through Monte Carlo simulation and study the network performance as a function of key constellation design parameters. We also analyse the impact of LEO SAT constellation parameters on the coverage capability and the average delay of MUs in this edge caching network. The simulation and analysis results provide constructive insights into the network design.

关键词： Geometry Performance evaluation Satellites Low earth orbit satellites Stochastic processes Quality of service Downlink Delays Numerical models Servers

来源：评论

学校读者我要写书评

暂无评论

Hybrid 2D/CMOS microchips for memristive applications

Hybrid 2D/CMOS microchips for memristive applications

引用

IEEE Nanotechnology Materials and Devices Conference (NMDC)

作者： Mario Lanza Materials Science and Engineering Program Physical Science and Engineering Division King Abdullah University of Science and Technology (KAUST) Thuwal Saudi Arabia

Two-dimensional layered materials (2D-LMs) materials have outstanding physical, chemical and thermal properties that make them attractive for the fabrication of solid-state micro/nano-electronic devices and circuits. However, synthesizing high-quality 2D-LMs at the wafer scale is difficult, and integrating them in semiconductor production lines brings associated multiple challenges. Nevertheless, in the past few years substantial progress has been achieved and leading companies like TSMC, Samsung and Imec have started to work more intensively on the fabrication of devices using 2D-LMs. In this invited talk, I will present our work towards hybrid 2D/CMOS microchips, with special emphasis on those dedicated to realize memristive operations. I will show state-of-the-art performance as electronic memory, as well as other novel properties that traditional microchips don’t exhibit, and that enable different innovative applications.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Back-end-of-line integration of 2D materials on silicon microchips

Back-end-of-line integration of 2D materials on silicon micr...

引用

International Electron Devices Meeting (IEDM)

作者： M. Lanza S. Pazos K. Zhu Y. Yuan Y. Shen O. Alharbi W. Zheng X. Zhang H. N. Alshareef Materials Science and Engineering Program Physical Science and Engineering Division King Abdullah University of Science and Technology (KAUST) Thuwal Saudi Arabia

Two-dimensional (2D) materials have excellent electronic and thermal properties that could help to improve the performance of electronic devices and circuits. However, scalable 2D materials synthesis and their integration in silicon microchips (with, for example, complementary metal-oxide-semiconductor, CMOS) transistors presents some challenges, mainly related to the presence of local defects that degrade device/circuit performance. In this invited article, we discuss the integration of 2D materials in silicon microchips. We analyze the type of materials and their properties, the main synthesis and manipulation methods, the type of circuits fabricated, the electronic performance achieved, as well as the future challenges and solutions. We focus on microchips with pre-patterned CMOS circuits on which the 2D material is deposited at the back-end-of-line (BEOL).

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Bi-population Cooperative Optimization Algorithm Assisted by an Autoencoder for Medium-scale Expensive Problems

引用

IEEE/CAA Journal of Automatica Sinica 2022年第11期9卷 1952-1966页

作者： Meiji Cui Li Li MengChu Zhou Jiankai Li abdullah Abusorrah Khaled Sedraoui IEEE the School of Intelligent Manufacturing Nanjing University of Science and TechnologyNanjing 210094China the Department of Electronics and Information Engineering Tongji UniversityShanghai 201804China the Helen and John C.Hartmann Department of Electrical and Computer Engineering New Jersey Institute of TechnologyNewarkNJ 07102 USA Center of Research Excellence in Renewable Energy and Power Systems Department of Electrical and Computer EngineeringFaculty of Engineeringand K.A.CARE Energy Research and Innovation CenterKing Abdulaziz UniversityJeddah 21589Saudi Arabia

This study presents an autoencoder-embedded optimization(AEO)algorithm which involves a bi-population cooperative strategy for medium-scale expensive problems(MEPs).A huge search space can be compressed to an informative lowdimensional space by using an autoencoder as a dimension reduction *** search operation conducted in this low space facilitates the population with fast convergence towards the *** strike the balance between exploration and exploitation during optimization,two phases of a tailored teaching-learning-based optimization(TTLBO)are adopted to coevolve solutions in a distributed fashion,wherein one is assisted by an autoencoder and the other undergoes a regular evolutionary ***,a dynamic size adjustment scheme according to problem dimension and evolutionary progress is proposed to promote information exchange between these two phases and accelerate evolutionary convergence *** proposed algorithm is validated by testing benchmark functions with dimensions varying from 50 to *** indicated in our experiments,TTLBO is suitable for dealing with medium-scale problems and thus incorporated into the AEO framework as a base *** with the state-of-the-art algorithms for MEPs,AEO shows extraordinarily high efficiency for these challenging problems,t hus opening new directions for various evolutionary algorithms under AEO to tackle MEPs and greatly advancing the field of medium-scale computationally expensive optimization.

关键词： Autoencoder dimension reduction evolutionary algorithm medium-scale expensive problems teaching-learning-based optimization

来源：评论

学校读者我要写书评

暂无评论

Newton Informed Neural Operator for Computing Multiple Solutions of Nonlinear Partials Differential Equations

arXiv

引用

arXiv 2024年

作者： Hao, Wenrui Liu, Xinliang Yang, Yahong Department of Mathematics The Pennsylvania State University University Park State CollegePA United States Computer Electrical and Mathematical Science and Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia

Solving nonlinear partial differential equations (PDEs) with multiple solutions using neural networks has found widespread applications in various fields such as physics, biology, and engineering. However, classical neural network methods for solving nonlinear PDEs, such as Physics-Informed Neural Networks (PINN), Deep Ritz methods, and DeepONet, often encounter challenges when confronted with the presence of multiple solutions inherent in the nonlinear problem. These methods may encounter ill-posedness issues. In this paper, we propose a novel approach called the Newton Informed Neural Operator, which builds upon existing neural network techniques to tackle nonlinearities. Our method combines classical Newton methods, addressing well-posed problems, and efficiently learns multiple solutions in a single learning process while requiring fewer supervised data points compared to existing neural network methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Newton-Raphson method

来源：评论

学校读者我要写书评

暂无评论

QoS Prediction Model of Cloud Services Based on Deep Learning

引用

IEEE/CAA Journal of Automatica Sinica 2022年第3期9卷 564-566页

作者： WenJun Huang PeiYun Zhang YuTong Chen MengChu Zhou Yusuf Al-Turki abdullah Abusorrah Nanjing University of Information Science&Technology School of Computer and SoftwareNanjingChina IEEE New Jersey Institute of Technology Helen and John C.Hartmann Department of Electrical and Computer EngineeringNewarkUSA Center of Research Excellence in Renewable Energy and Power Systems Department of Electrical and Computer EngineeringFaculty of Engineeringand K.A.CARE Energy Research and Innovation CenterKing Abdulaziz UniversityJeddah 21589Saudi Arabia

Dear editor,This letter presents a deep learning-based prediction model for the quality-of-service(QoS)of cloud ***,to improve the QoS prediction accuracy of cloud services,a new QoS prediction model is proposed,which is based on multi-staged multi-metric feature fusion with individual *** multi-metric features include global,local,and individual *** results show that the proposed model can provide more accurate QoS prediction results of cloud services than several state-of-the-art methods.

关键词： services Services individual

来源：评论

学校读者我要写书评

暂无评论

Flexible Hall sensor made of laser-scribed graphene

引用

npj Flexible Electronics 2021年第1期5卷 18-24页

作者： Altynay Kaidarova Wenhao Liu Liam Swanepoel abdullah Almansouri Nathan RGeraldi Carlos.M.Duarte Jurgen Kosel Computer Electrical and Mathematical Sciences&Engineering(CEMSE)King Abdullah University of Science and Technology(KAUST)ThuwalSaudi Arabia Department of Electrical and Electronic Engineering University of JeddahDhahbanSaudi Arabia Red Sea Research Center(RSRC)and Computational Biosciences Research Center King Abdullah University of Science and Technology(KAUST)ThuwalSaudi Arabia Sensor Systems Division Silicon Austria Labs(SAL)VillachAustria

Graphene has shown considerable potential for sensing magnetic fields based on the Hall Effect,due to its high carrier mobility,low sheet carrier density,and low-temperature ***,the cost of graphene in comparison to conventional materials has meant that its uptake in electronic manufacturing has been *** lower technological barriers and bring more widespread adoption of graphene Hall sensors,we are using a one-step laser scribing process that does not rely on multiple steps,toxic chemicals,and subsequent ***-scribed graphene Hall sensors offer a linear response to magnetic fields with a normalized sensitivity of~1.12 V/*** also exhibit a low constant noise voltage floor of~50 nV√Hz p for a bias current of 100μA at room temperature,which is comparable with state-of-the-art low-noise Hall *** sensors combine a high bendability,come with high robustness and operating temperatures up to 400°*** enable device ideas in various areas,for instance,soft *** an example,we combined a laser-scribed graphene sensor with a deformable elastomer and flexible magnet to realize low-cost,compliant,and customizable tactile sensors.

关键词： normalized instance steps

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：