检索结果-内蒙古大学图书馆

The Journal of Machine Learning Research 2020年第1期21卷 4508-4531页

作者： Houssem Sifaou Abla Kammoun Mohamed-Slim Alouini Computer Electrical and Mathematical Science and Engineering Division King Abdullah University of Science and Technology Thuwal KSA

Linear discriminant analysis (LDA) is a popular classifier that is built on the assumption of common population covariance matrix across classes. The performance of LDA depends heavily on the quality of estimating the mean vectors and the population covariance matrix. This issue becomes more challenging in high-dimensional settings where the number of features is of the same order as the number of training samples. Several techniques for estimating the covariance matrix can be found in the literature. One of the most popular approaches are estimators based on using a regularized sample covariance matrix, giving the name regularized LDA (R-LDA) to the corresponding classifier. These estimators are known to be more resilient to the sample noise than the traditional sample covariance matrix estimator. However, the main challenge of the regularization approach is the choice of the optimal regularization parameter, as an arbitrary choice could lead to severe degradation of the classifier performance. In this work, we propose an improved LDA classifier based on the assumption that the covariance matrix follows a spiked covariance model. The main principle of our proposed technique is the design of a parametrized inverse covariance matrix estimator, the parameters of which are shown to be easily optimized. Numerical simulations, using both real and synthetic data, show that the proposed classifier yields better classification performance than the classical R-LDA while requiring lower computational complexity.

关键词： linear discriminant analysis spiked covariance models high-dimensional data random matrix theory

来源：评论

学校读者我要写书评

暂无评论

Contextual Multi-Armed Bandit based Beam Allocation in mmWave V2X Communication under Blockage

Contextual Multi-Armed Bandit based Beam Allocation in mmWav...

引用

IEEE Conference on Vehicular technology (VTC)

作者： Arturo Medina Cassillas Abdulkadir Kose Haeyoung Lee Chuan Heng Foh Chee Yen Leow Department of Engineering King’s College London London UK Department of Computer Engineering Abdullah Gul University Kayseri Turkey School of Physics Engineering and Computer Science University of Hertfordshire Hatfield UK 5GIC & 6GIC Institute for Communication Systems (ICS) University of Surrey Guildford UK Wireless Communication Centre Faculty of Electrical Engineering Universiti Teknologi Malaysia Johor Malaysia

Due to its low latency and high data rates support, mmWave communication has been an important player for vehicular communication. However, this carries some disadvantages such as lower transmission distances and inability to transmit through obstacles. This work presents a Contextual Multi-Armed Bandit Algorithm based beam selection to improve connection stability in next generation communications for vehicular networks. The algorithm, through machine learning (ML), learns about the mobility contexts of the vehicles (location and route) and helps the base station make decisions on which of its beam sectors will provide connection to a vehicle. In addition, the proposed algorithm also smartly extends, via relay vehicles, beam coverage to outage vehicles which are either in NLOS condition due to blockages or not served any available beam. Through a set of experiments on the city map, the effectiveness of the algorithm is demonstrated, and the best possible solution is presented.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Dual non-diffractive terahertz beam generators based on all-dielectric metasurface

引用

Frontiers of Optoelectronics 2021年第2期14卷 201-210页

作者： Chunyu LIU Yanfeng LI Xi FENG Xixiang ZHANG Jiaguang HAN Weili ZHANG Center for Terahertz Waves College of Precision Instrument and Optoelectronics EngineeringKey Laboratory of Optoelectronics Information and Technology(Ministry of Education of China)Tianjin UniversityTianjin 300072China Physical Science and Engineering Division King Abdullah University of Science and TechnologyThuwal 23955-6900Saudi Arabia School of Electrical and Computer Engineering Oklahoma State UniversityStillwaterOK 74078USA

The applications of terahertz(THz)technology can be greatly extended using non-diffractive beams with unique field distributions and non-diffiactive transmission ***,we design and experimentally demonstrate a set of dual non-diffractive THz beam generators based on an all-dielectric *** kinds of non-diffractive beams with dramatically opposite focusing properties,Bessel beam and abruptly autofocus-ing(AAF)beam,are considered.A Bessel beam with longdistance non-diffractive characteristics and an AAF beam with low energy during transmission and abruptly increased energy near the focus are generated for x-and^-polarized incident waves,*** two kinds of beams are characterized and the results agree well with *** addition,we show numerically that these two kinds of beams can also carry orbital angular momentum by further imposing proper angular phases in the *** believe that these metasurface-based beam generators have great potential use in THz imaging,communications,non-destructive evaluation,and many other fields.

关键词： terahertz(THz)wave all-dielectric metasurface Bessel beam abruptly autofocusing(AAF)beam vortex beam

来源：评论

学校读者我要写书评

暂无评论

Boundary stabilization of a reaction-diffusion system weakly coupled at the boundary.

引用

IFAC-PapersOnLine 2020年第2期53卷 16537-16542页

作者： Mohamed Ghattassi Taous Meriem Laleg Computer Electrical and Mathematical Science and Engineering (CEMSE) Division King Abdullah university of science and technology KSA

This study analyses the boundary stabilization of a system of two parabolic linear PDEs weakly coupled at the boundary. This model is motivated by heat transfer in a membrane distillation based desalination modeled by a two-dimensional advection diffusion equations coupled at the boundary. Based on some physical assumptions, the 2D model can be formulated as a 1D reaction-diffusion system. Two cases were studied: full and under actuated scenarios. In the full actuated case, a backstepping approach is used to map the plant to an exponentially stable target system. The well-posedness of the kernel equations is proved. Moreover, the actuation of only one of the parabolic equations has been considered. The standard backstepping transformations is again used to transform the initial plant to a desired target system where Lyapunov analysis is adequately used. Finally, a numerical example showing the performance of the proposed control design is presented.

关键词： Backstepping reaction-diffusion system boundary Stabilization Membrane Distillation full/under actuated

来源：评论

学校读者我要写书评

暂无评论

Unlocking Complex Chemical and Morphological Transformations During Thermal Treatment of O-Hydroxyl-Substituted Polyimide of Intrinsic Microporosity: Impact on Ethanol/Cyclohexane Separation

SSRN

引用

SSRN 2023年

作者： Ogieglo, Wojciech Knozowska, Katarzyna Puspasari, Tiara Ali, Zain Kujawa, Joanna Korczeniewski, Emil Terzyk, Artur P. Kujawski, Wojciech Pinnau, Ingo Functional Polymer Membranes Group Chemical Engineering Program Division of Physical Science and Engineering King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia Advanced Membranes and Porous Materials Center Division of Physical Science and Engineering King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia Nicolaus Copernicus University in Toruń Membranes and Membrane Techniques Research Group Faculty of Chemistry 7 Gagarina Street Toruń87-100 Poland Nicolaus Copernicus University Faculty of Chemistry Physicochemistry of Carbon Materials Research Group Faculty of Chemistry Gagarin Street 7 Toruń87-100 Poland

In this work, we investigated the pervaporation separation performance of an ethanol – cyclohexane mixture using a thin-film composite membrane based on a heat-treated, o-hydroxyl-substituted polyimide of intrinsic microporosity (PIM-6FDA-OH). We report in detail the complex chemical and structural transformations that are associated with the development of thermally-rearranged (TR, 400 – 450 °C), and carbon molecular sieve (CMS) membranes (> 500 °C). These transformations strongly affect the affinity-related (e.g. swelling behavior) and morphology-related (e.g. microporosity) properties of the TR and CMS membrane materials which translate into a complex molecular-level separation behavior toward an organic solvent mixture during pervaporation. We show that excellent separation performance (total flux ~ 4 kg / m2h and ethanol/cyclohexane separation factor of ~1250) can be achieved after precise thermal treatment at the onset of the CMS formation. In addition, we highlight the unique property of pervaporation where, in contrast to e.g. liquid solvent permeation, the downstream side of the membrane remains essentially dry due to exposure to high vacuum. As a result, a sharp molecular cut-off can be achieved within this dry downstream region of the membrane despite considerable overall swelling/penetrant concentration in the upstream region of the membrane. © 2023, The Authors. All rights reserved.

关键词： Pervaporation

来源：评论

学校读者我要写书评

暂无评论

Signal Processing and Machine Learning Techniques for Terahertz Sensing: An Overview

arXiv

引用

arXiv 2021年

作者： Helal, Sara Sarieddeen, Hadi Dahrouj, Hayssam Al-Naffouri, Tareq Y. Alouini, Mohamed-Slim The Department of Electrical Engineering Effat University Jeddah22332 Saudi Arabia The Division of Computer Electrical and Mathematical Sciences and Engineering King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia The Center of Excellence for NEOM Research King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia

Following the recent progress in Terahertz (THz) signal generation and radiation methods, joint THz communications and sensing applications are being proposed for future wireless systems. Towards this end, THz spectroscopy is expected to be carried over user equipment devices to identify material and gaseous components of interest. THz-specific signal processing techniques should complement this re-surged interest in THz sensing for efficient utilization of the THz band. In this paper, we present an overview of these techniques, with an emphasis on signal pre-processing (standard normal variate normalization, min-max normalization, and Savitzky-Golay filtering), feature extraction (principal component analysis, partial least squares, t-distributed stochastic neighbor embedding, and nonnegative matrix factorization), and classification techniques (support vector machines, k-nearest neighbor, discriminant analysis, and naive Bayes). We also address the effectiveness of deep learning techniques by exploring their promising sensing and localization capabilities at the THz band. Lastly, we investigate the performance and complexity trade-offs of the studied methods in the context of joint communications and sensing (JCAS). We thereby motivate the corresponding use-cases, and present a handful of contextual future research directions. Copyright © 2021, The Authors. All rights reserved.

关键词： Terahertz spectroscopy

来源：评论

学校读者我要写书评

暂无评论

Utilizing machine learning algorithms for task allocation in distributed agile software development

引用

Heliyon 2024年第21期10卷 e39926页

作者： Al-Fraihat, Dimah Sharrab, Yousef Al-Ghuwairi, Abdel-Rahman Alzabut, Hamza Beshara, Malik Algarni, Abdulmohsen Department of Software Engineering Faculty of Information Technology Isra University Amman 11622 Jordan Department of Data Science and Artificial Intelligence Faculty of Information Technology Isra University Amman 11622 Jordan Department of Software Engineering Faculty of Prince Al-Hussien bin Abdullah for IT The Hashemite University Zarqa 13133 Jordan Department of Computer Science King Khalid University Abha 61421 Saudi Arabia

Distributed agile software development (DASD) has become a prominent software development approach. Proper task allocation is crucial in DASD to avoid undesirable outcomes including project rejection by clients, unfavorable team attitudes, and project failure. Coordination and communication issues occur as businesses embrace the DASD environment more frequently to tap into global talent and knowledge while cutting development expenses. To overcome these challenges, efficient task allocation planning becomes a crucial success component in software project management. The purpose of this study is to utilize machine learning (ML) predictive algorithms to determine the most appropriate role for a given task, with the aim of assisting software managers in making task assignments more efficiently and effectively in DASD environment. Preprocessing steps applied to the dataset include data cleaning, normalization, and partitioning into training, validation, and test sets. Four model classifiers were used in the experiment: Random Forest, Decision Tree, K-Nearest Neighbors (K-NN), and AdaBoost. The results showed that Random Forest outperformed the other classifiers in task allocation prediction, achieving an accuracy of 96.7 %, followed by K-NN (94.2 %), Decision Tree (93.5 %), and AdaBoost (93 %). The study demonstrates that ML models are effective in tackling task allocation issues in DASD settings, and the outcomes are promising. © 2024

关键词： DASD Distributed agile software development Machine learning (ML) Software engineering Software project management Task allocation

来源：评论

学校读者我要写书评

暂无评论

Topology Rewiring Strategies to construct robust scale-free medical Internet of Things Networks

引用

Simulation Modelling Practice and Theory 2025年 142卷

作者： Khan, Muhammad Awais Javaid, Nadeem Alrajeh, Nabil Bouk, Safdar Hussain Department of Electrical and Computer Engineering COMSATS University Islamabad Islamabad44000 Pakistan ComSens Lab International Graduate School of Artificial Intelligence National Yunlin University of Science and Technology Yunlin64002 Taiwan Department of Biomedical Technology College of Applied Medical Sciences King Saud University Riyadh11633 Saudi Arabia Old Dominion University 5115 Hampton Blvd NorfolkVA23529 United States

The Internet of Things (IoT) network topologies are now most commonly impacted by cyberattacks. The scale-free network topologies have demonstrated great robustness against random attacks by preserving the connectedness of the nodes. The scale-free network topologies’ susceptibility to malicious attacks, however, is a significant worry. It is due to the significance of the scale-free networks in different fields of life like medical, transportation, education, agriculture, etc. Also, high-degree node removal diminishes the network's resiliency and compromises the connection of the majority of nodes. In this study, we offer several rewiring techniques for building scale-free, reliable Medical Internet of Things (MIoT) networks that can withstand malicious attacks. Initially, the scale-free MIoT network's performance optimization is ensured using a heuristic algorithm known as the Great Deluge Algorithm (GDA). Then, four rewiring strategies are formulated. The initial approach is degree dissortativity, which rewires the network if all nodes have high maximum connectivity to other neighbors with a similar degree. For the second strategy, we introduced a degree difference operation based on degree dissortativity to ensure that the edges that are connected possess low dissortativity and degree difference. Meanwhile, the remaining two strategies take into account the node load bound and enhanced GDA to increase network robustness. The performance of the proposed rewiring strategies is validated through simulations. The results prove that the proposed strategies increase network robustness by up to 25% compared to Hill Climbing (HC) and Simulated Annealing (SA). Additionally, the strategies show great success in improving network connectivity and graph density. However, their computational time is higher compared to HC and SA. © 2025 Elsevier B.V.

关键词： Network topology

来源：评论

学校读者我要写书评

暂无评论

Diversity Schemes in Multi-hop Visible Light Communications for 6G Networks

引用

Procedia computer science 2021年 182卷 140-149页

作者： Alice Faisal Rawan Alghamdi Hayssam Dahrouj Hadi Sarieddeen Nasir Saeed Tareq Y. Al-Naffouri Mohamed-Slim Alouini Faculty of Engineering and Applied Science Memorial University of Newfoundland St. John\x92s NL A1C 5S7 Canada Division of Computer Electrical and Mathematical Sciences and Engineering King Abdullah University of Science and Technology Thuwal 23955-6900 Saudi Arabia Center of Excellence for NEOM Research Division of Computer Electrical and Mathematical Sciences and Engineering King Abdullah University of Science and Technology Thuwal 23955-6900 Saudi Arabia National University of Technology Islamabad Pakistan

With the fast-emerging trend of deploying Internet-of-things (IoT) in smart homes and cities, next-generation wireless systems (i.e., beyond fifth-generation (5G) of wireless mobile communications) are expected to connect a plurality of devices, including machines, sensors, tablets, cameras, etc. To this end, the radio frequency (RF) spectrum would fail to fulfill the demands of such enormous data-heavy wireless applications, as operating at the RF spectrum would largely suffer from data congestion and interference. Based on the potential of Visible Light Communications (VLC) to solve the problem of spectrum scarcity and satisfy the high data-rate requirements of beyond 5G systems, this paper focuses on utilizing VLC for IoT applications in indoor environments by means of adequately routing data among VLCbased nodes. Operating at the unlicensed optical band in the visible light region utilizes simple Light-Emitting-Diodes (LEDs), which is a health-friendly communication method. VLC has the advantages of providing ultra-high bandwidth, robustness to electromagnetic interference, and inherent physical security. VLC, however, suffers from severe short communication ranges and line-of-sight (LoS) constraints. This paper proposes overcoming such challenges by means of adopting a transmit diversity scheme that transmits messages over several paths. In order to reduce the receiver outage probability, the paper proposes multiple combination schemes at the receiver side, namely selection combining, maximal-ratio combining, and threshold combining. The simulation results assess the effectiveness of the proposed diversity schemes in terms of signal-to-noise-ratio and outage probability and illustrate the suitability of deploying VLC for next-generation indoor networks. The results show that that maximal-ratio combining outperforms selection combining and threshold combining by approximately 10% and 40%, respectively.

关键词： 6G Visible Light Communication Diversity Combination Protocols Internet-of-things Indoor Applications

来源：评论

学校读者我要写书评

暂无评论

Momentum-based ICA for Self Interference Cancellation in In-Band Full-Duplex Systems

Momentum-based ICA for Self Interference Cancellation in In-...

引用

Asilomar Conference on Signals, Systems & computers

作者： Chi Lee Chung-An Shen Mohammed E. Fouda Ahmed Eltawil Electronic and Computer Engineering Dept. National Taiwan University of Science and Technology Taipei Taiwan Center for Embedded & Cyber-physical Systems University of California-Irvine Irvine CA USA King Abdullah University of Science and Technology Thuwal Saudi Arabia

Recently, Independent Component Analysis (ICA) has proven its effectiveness as a self-interference cancellation method for in-band full duplex systems. However, ICA could suffer slow convergence due to the iterative estimation of the independent components which limits its usage in real-time applications. In this paper, we introduce a momentum-based ICA to accelerate convergence via incorporating gradient history. The proposed momentum-based ICA is evaluated and tested on different ICA algorithms including real-valued and complexvalued FastICA and entropy bound minimization based ICA. The results show significant speedup improvement compared to native ICA based on gradient descent approach. The proposed algorithm shows consistent results under different transceiver non-linearity and for different frame lengths.

关键词： Interference cancellation Estimation Full-duplex system Independent component analysis Minimization Transceivers Real-time systems

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：