检索结果-内蒙古大学图书馆

SSRN 2022年

作者： Dorozhovets, Mykhaylo Pawlowski, Eligiusz Świsulski, Dariusz Department of Metrology and Diagnostic Systems Faculty of Electrical and Computer Engineering Rzeszow University of Technology ul. Wincentego Pola 2A Rzeszów35-959 Poland Department of Information and Measuring Technology National University Lviv Polytechnic Lviv Ukraine Faculty of Electrical Engineering and Computer Science Lublin University of Technology Lublin Poland Faculty of Electrical and Control Engineering Gdańsk University of Technology Gdańsk Poland

The paper presents investigation of the efficiency of weight averaging of a pulse frequency modulated signal. The sampling effects and the interference influence in the input signal of the voltage-to frequency converter on the result of output frequency weight averaging are analyzed. A general form of the averaging result dependence on the used weighting function, signal value, interference level and modulation parameters is derived. In the case of high interferences whose frequencies are unstable or may change over a wide range, the specified level of their suppression, together with the reduction of the sampling effect, can be achieved using weighting functions, such as trigonometric, with an appropriate level of side lobes. The interference suppression dependently on frequency ranges using selected weighing functions are tested using both the simulation and experimental methods. The obtained results show very good convergence with the values calculated from theoretical formulas. © 2022, The Authors. All rights reserved.

关键词： Weighing

来源：评论

学校读者我要写书评

暂无评论

PR-DARTS: Pruning-Based Differentiable Architecture Search

SSRN

引用

SSRN 2022年

作者： Mousavi, Seyedhamidreza Hamid Loni, Mohammad Alibeigi, Mina Daneshtalab, Masoud School of Innovation Design and Engineering Malardalen University Vasteras72218 Sweden Zenseact AB Lindholmspiren2 Goteborg41756 Sweden The Department of Computer Science and Technology The University of Cambridge 15 JJ Thomson Ave CambridgeCB3 0FD United Kingdom School of Innovation Design and Engineering Malardalen Vasteras72218 Sweden Department of Computer Systems TalTech University Tallinn19086 Estonia

The deployment of Convolutional Neural Networks (CNNs) on edge devices is hindered by the substantial gap between performance requirements and available processing power. While recent research has made large strides in developing network pruning methods for reducing the computing overhead of CNNs, there remains considerable accuracy loss, especially at high pruning ratios. We find that the architectures designed for non-pruned networks are not effective for pruned networks since there is no support for characteristics of pruned operations in the search space. Plus, conventional NAS methods use a search objective specialized for non-pruned networks. We propose an algorithm to search architectures for pruning methods by defining a new search space and a novel search objective. To improve the generalization of the pruned networks, we propose two novel PrunedConv and PrunedLinear operations. Specifically, these operations mitigate the problem of unstable gradients by regularizing the objective function of the pruned networks. The proposed search objective enables us to train architecture parameters regarding the pruned weight elements. Quantitative analyses demonstrate that our searched architectures outperform those used in the state-of-the-art pruning networks on CIFAR-10 and ImageNet datasets. In terms of hardware effectiveness, PR-DARTS increases the accuracy of the pruned version of MobileNet-v2 from 73.44% to 81.35% (+7.91% improvement) with 3.87× faster inference time. © 2022, The Authors. All rights reserved.

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

Mirror Descent and Novel Exponentiated Gradient Algorithms Using Trace-Form Entropies and Deformed Logarithms

arXiv

引用

arXiv 2025年

作者： Cichocki, Andrzej Tanaka, Toshihisa Cruces, Sergio Systems Research Institute of Polish Academy of Science Newelska 6 Warszawa01-447 Poland Gagarina 11 87-100Torun Poland Global Innovation Research Institute Tokyo University of Agriculture and Technology 2-24-16 Naka-cho Koganei-shi Tokyo Japan Riken AIP Nihonbashi 1 chome-4-1 Tokyo103-0027 Japan Department of Electrical Engineering and Computer Science Tokyo University of Agriculture and Technology 2-24-16 Naka-cho Koganei-shi Tokyo Japan Departamento de Teoria de la Senal y Comunicaciones Universidad de Sevilla Camino de los Descubrimientos s/n Seville41092 Spain

In this paper we propose and investigate a wide class of Mirror Descent updates (MD) and associated novel Generalized Exponentiated Gradient (GEG) algorithms by exploiting various trace-form entropies and associated deformed logarithms and their inverses - deformed (generalized) exponential functions. The proposed algorithms can be considered as extension of entropic MD and generalization of multiplicative gradient updates. In the literature, there exist nowadays over fifty mathematically well defined generalized entropies, so impossible to exploit all of them in one research paper. So, we focus only on several selected most popular entropies and associated logarithms like the Tsallis, Kaniadakis and Sharma-Taneja-Mittal and some of their extension like Tempesta or Kaniadakis-Scarfone entropies. The shape and properties of the deformed logarithms and their inverses are adjusted by one or more hyperparameters. By learning these hyperparameters, we can adapt to distribution of training data, which can be designed to the specific geometry of the optimization problem, leading to potentially faster convergence and better performance. The using generalized entropies and associated deformed logarithms in the Bregman divergence, used as a regularization term, provides some new insight into mirror descent and exponentiated gradient descent updates. © 2025, CC BY-NC-ND.

关键词： Exponential functions

来源：评论

学校读者我要写书评

暂无评论

Privacy-preserving Intelligent Resource Allocation for Federated Edge Learning in Quantum Internet

arXiv

引用

arXiv 2022年

作者： Xu, Minrui Niyato, Dusit Yang, Zhaohui Xiong, Zehui Kang, Jiawen Kim, Dong In Shen, Xuemin The School of Computer Science and Engineering Nanyang Technological University Singapore The College of Information Science and Electronic Engineering Zhejiang University Hangzhou310007 China Hangzhou310007 China Zhejiang Laboratory Hangzhou31121 China The Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore 487372 Singapore The School of Automation Guangdong University of Technology China Department of Electrical and Computer Engineering Sungkyunkwan University Suwon16419 Korea Republic of The Department of Electrical and Computer Engineering University of Waterloo Waterloo ON N2L 3G1 Canada

Federated edge learning (FEL) is a promising paradigm of distributed machine learning that can preserve data privacy while training the global model collaboratively. However, FEL is still facing model confidentiality issues due to eavesdropping risks of exchanging cryptographic keys through traditional encryption schemes. Therefore, in this paper, we propose a hierarchical architecture for quantum-secured FEL systems with ideal security based on the quantum key distribution (QKD) to facilitate public key and model encryption against eavesdropping attacks. Specifically, we propose a stochastic resource allocation model for efficient QKD to encrypt FEL keys and models. In FEL systems, remote FEL workers are connected to cluster heads via quantum-secured channels to train an aggregated global model collaboratively. However, due to the unpredictable number of workers at each location, the demand for secret-key rates to support secure model transmission to the server is unpredictable. The proposed systems need to efficiently allocate limited QKD resources (i.e., wavelengths) such that the total cost is minimized in the presence of stochastic demand by formulating the optimization problem for the proposed architecture as a stochastic programming model. To this end, we propose a federated reinforcement learning-based resource allocation scheme to solve the proposed model without complete state information. The proposed scheme enables QKD managers and controllers to train a global QKD resource allocation policy while keeping their private experiences local. Numerical results demonstrate that the proposed schemes can successfully achieve the cost-minimizing objective under uncertain demand while improving the training efficiency by about 50% compared to state-of-the-art schemes. © 2022, CC BY.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Construction of Hunt Processes by the Lyapunov Method and Applications to Generalized Mehler Semigroups

SSRN

引用

SSRN 2025年

作者： Cîmpean, Iulian Beznea, Lucian Röckner, Michael Simion Stoilow Institute Mathematics of the Romanian Academy Research unit No. 2 P.O. Box 1-764 BucharestRO-014700 Romania POLITEHNICA Bucharest CAMPUS Institute Romania Department of Mathematics Faculty of Mathematics and Computer Science University of Bucharest 14 Academiei Bucharest010014 Romania Simion Stoilow Institute Mathematics of the Romanian Academy 21 Calea Grivit ¸ei Bucharest010702 Romania Fakultät für Mathematik Universität Bielefeld Postfach 100 131 BielefeldD-33501 Germany Academy for Mathematics and Systems Science CAS Beijing China

In this paper we deal with the problem of characterizing those generalized Mehler semigroups that correspond to càdlàg Markov processes, which has remained open for more than a decade. Our approach is to reconsider the càdlàg problem for generalized Mehler semigroups as a particular case of the much broader problem of constructing Hunt (hence càdlàg and quasi-left continuous) processes from a given Markov semigroup. Following this strategy, we derive checkable sufficient conditions for a large class of generalized Mehler semigroups in order to posses an associated Hunt process with values in the original space, in contrast to previous results where an extension of the state space was required. We apply the general results to solve a stochastic heat equation on $L^2(D)$ whose drift is given by the Dirichlet Laplacian on a bounded domain $D \subset \mathbb{R}^d$, driven by a (non-diagonal) Lévy noise whose characteristic exponent is not necessarily Sazonov continuous. © 2025, The Authors. All rights reserved.

关键词： Lyapunov functions

来源：评论

学校读者我要写书评

暂无评论

CARDIOCARE platform: A beyond the state of the art approach for the management of elderly multimorbid patients with breast cancer therapy induced cardiac toxicity

CARDIOCARE platform: A beyond the state of the art approach ...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Tsiouris, Kostas M. Kalliatakis, Grigoris Mazzocco, Ketti Seruga, Bostjan Marias, Kostas Karanasiou, Georgia Antoniades, Athos Papakonstantinou, Andri Conti, Constanza Tsiknakis, Manolis Sfakianakis, Stelios Lakkas, Lampros Filippatos, Gerasimos Bucur, Anca Fotiadis, Dimitrios I. Manikis, George Mauri, Davide Constantinidou, Anastasia Pacella, Elsa University of Ioannina Unit of Medical Technology and Intelligent Information Systems Ioannina Greece Foundation for Research and Technology - Hellas Computational Bio-Medicine Laboratory Institute of Computer Science Heraklion Greece European Institute of Oncology Cardiology Division Milan Italy Institute of Oncology Ljubljana Department of Medical Oncology Ljubljana Slovenia Stremble Ventures Ltd Limassol Cyprus Karolinska University Hospital Endocrine Tumors and Sarcoma Theme Cancer Dept. for Breast Cancer Sweden Istituto di Management Sanitario S.r.l Milano Italy Foundation for Research and Technology - Hellas Institute of Computer Science Heraklion Greece University of Ioannina 2nd Dept. of Cardiology School of Medicine Ioannina Greece National and Kapodistrian University of Athens Athens Greece Attikon University Hospital School of Medicine Dept. of Cardiology Athens Greece Philips Research Europe Eindhoven Netherlands University of Ioannina Dept. of Medical Oncology Ioannina Greece Bank of Cyprus Oncology Centre Nicosia Cyprus European Society of Cardiology Brussels Belgium

ISBN: (纸本)9798350337488

Breast cancer (BC) is the most common cancer in women in Europe and worldwide, with a high prevalence in middle-aged and older women. The last years, the evolution in the existing treatment approaches have contributed to improved clinical outcomes and survival rates. Nevertheless, BC therapy-related cardiotoxicity, poses a severe impact in the short- and long-term Quality of Life (QoL) and associated survival of the BC patients. This study demonstrates how the CARDIOCARE platform and the developed risk stratification models provides healthcare professionals with a valuable tool for effectively managing BC patients, preventing treatment induced cardiotoxicity and improving their QoL. This is accomplished through the integration of multi-source patient-specific data from patient-oriented mobile applications and wearable sensors, and by the employment of beyond the state-of-the-art data mining and machine learning approaches. © 2023 IEEE.

关键词： Breast cancer cardiotoxicity machine learning mobile application risk stratification models

来源：评论

学校读者我要写书评

暂无评论

Performance evaluation for ranking-based discretisation

引用

Procedia computer science 2020年 176卷 3335-3344页

作者： Grzegorz Baron Urszula Stańczyk Department of Graphics Computer Vision and Digital Systems Faculty of Automatic Control Electronics and Computer Science Silesian University of Technology Akademicka 2A 44-100 Gliwice Poland

Discretisation often constitutes a part of initial data preparation stage. It translates continuous domain of features into granular, by assigning a number of intervals to represent attributes’ values by nominal categories. Typically all real-valued features are subjected to transformations, regardless of their characteristics. The paper presents research on discretisation executed with a discerning approach. To all available attributes, feature selection mechanisms were employed, in the form of rankings that order variables based on their importance. Exploiting this discovered knowledge on attributes, discretisation was then driven by a ranking, and either highest or lowest ranking features were selected for transformation. The influence of selective discretisation on the performance of classification systems was studied for three popular inducers. The procedure was employed in the field of stylometry, and a task of authorship recognition, considered as a binary classification with balanced classes. The experiments show that discretisation based on importance of features can lead to better performance than in the case of transformations applied to all attributes.

关键词： discretisation feature selection ranking importance of attributes stylometry classification

来源：评论

学校读者我要写书评

暂无评论

Service Migrations in TSCH Network using Wireless Channel Estimation and Prediction 50

Service Migrations in TSCH Network using Wireless Channel Es...

引用

50. Jahrestagung der Gesellschaft fur Informatik, INFORMATIK 2020 - 50th Annual Conference of the German Informatics Society, INFORMATIK 2020

作者： Nikoukar, Ali Raza, Saleem Rao, Tharakeswara Güneş, Mesut Dezfouli, Behnam Otto-von-Guericke Universitčt Communication and Networked Systems Universitčtsplatz 2 Magdeburg39106 Germany Internet of Things Research Lab Department of Computer Science and Engineering Santa Clara University Santa Clara United States

ISBN: (纸本)9783885797012

Industrial Wireless Sensors and Actuators Networks (IWSANs) are gateway to the Industrial 4.0, which promises to realize smart factory leading to the Industrial Internet of Things (IIoT). It employs Cyber-Physical systems (CPSs) to enhance operational efficiency and flexibility while reducing cost. IWSANs are delay-sensitive and always require low latency and reliable connection from sensor to actuator to successfully perform a physical action. Reliability and low-latency complement each other to prevent expected failures in wireless medium. In this way, detecting and predicting failure before it actually occurs is key to actually avoid it well in time. Detection and predictions are imperative in locating faults and failures. The causes of failures in a sensor or actuator can include hardware malfunction, poor battery life, interference, accident, and short term wireless connectivity problems. Although, industrial environment mostly undertakes redundant resource to circumvent such issues, yet poor coordination among multiple resources and inaccurately predicting failures may result in losses. In such a scenario, migration of services come to be a rescue, where an intermediary can migrate service from one device, which cannot complete a task due to resource exhaustion, to a more resource-rich device. Thus, in this paper, we focus on wireless connectivity failures caused by interference in the 2.4GHz frequency band. We do it by designing an Multi Channel Sniffing Setup (MCSS) testbed, that acts as a spectrum observer and is deployed in different locations in industrial WSAN. Alongside, we use the concept of Cognitive Radio (CR) to predict interference and noise level in the spectrum by proposing an Intelligent Low-power Wireless Spectrum Prediction (ILPWSP) based on Deep Q Network (DQN). The MCSS testbed and the ILPWSP coordinate in assessing wireless connectivity risks, predict failures in sensor and actuator nodes and then make efficient decisions on the migration o

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Dog Noseprint Identification Based on Textural Features of Scaly Blocks

SSRN

引用

SSRN 2022年

作者： Chan, Yung-Kuan Lin, Chuen-Horng Wang, Ching-Lin Tu, Keng-Chang Yang, Shu-Chun Tsai, Meng-Hsiun Yug, Shyr-Shen Department of Management Information Systems National Chung Hsing University 145 Xingda Rd. South Dist. Taichung City402 Taiwan Department of Computer Science and Engineering National Chung Hsing University 145 Xingda Rd. South Dist. Taichung City402 Taiwan Department of Computer Science and Information Engineering National Taichung University of Science and Technology No. 129 Sec. 3 Sanmin Rd Taichung Taiwan Department of Information Management National Chin-Yi University of Technology No.57 Sec. 2 Zhongshan Rd. Taiping Dist. Taichung411030 Taiwan

Because dog noseprints are equivalent to human fingerprints and have unique characteristics, this study proposes a dog identification technology based on dog noseprints for identifying and managing stray animals. This study presents three processing stages. The first stage uses YOLOv3 to detect the dog's nose and the nostril regions. The second stage involves enhancing the contrast of the image and the contour of the scaly block using the multi-scale line detector. The third stage is to extract the shape and spatial features of the scaly block and use them to identify dogs. This study included 276 dogs from multiple animal families and public shelters in Taiwan. The dataset was randomly divided into three groups to determine the best parameters for matching the dog noseprint via the experiment. The accuracy (ACC) of each dog identification group exceeded 97.83%, thereby proving that the parameter-matching method proposed herein offers high stability. In an additional experimental dataset of dog noseprint images used to perform dog identification, the ACC of the proposed method exceeded 90.22% in the top 1 and 94.57% in the top 3. In addition, the ACC results of the different groups demonstrate that the proposed method can consistently obtain high accuracy in dog identification. © 2022, The Authors. All rights reserved.

关键词： Animals

来源：评论

学校读者我要写书评

暂无评论

Ornstein-Uhlenbeck process and generalizations: particle’s dynamics under comb constraints and stochastic resetting

arXiv

引用

arXiv 2023年

作者： Trajanovski, Pece Jolakoski, Petar Zelenkovski, Kiril Iomin, Alexander Kocarev, Ljupco Sandev, Trifce Research Center for Computer Science and Information Technologies Macedonian Academy of Sciences and Arts Bul. Krste Misirkov 2 Skopje1000 Macedonia Department of Physics Technion Haifa32000 Israel Max Planck Institute for the Physics of Complex Systems Dresden Germany Faculty of Computer Science and Engineering Ss. Cyril and Methodius University PO Box 393 Skopje1000 Macedonia Institute of Physics & Astronomy University of Potsdam Potsdam-GolmD-14776 Germany Institute of Physics Faculty of Natural Sciences and Mathematics Ss. Cyril and Methodius University Arhimedova 3 Skopje1000 Macedonia

The Ornstein-Uhlenbeck process is interpreted as Brownian motion in a harmonic potential. This Gaussian Markov process has a bounded variance and admits a stationary probability distribution, in contrast to the standard Brownian motion. It also tends to a drift towards its mean function, and such a process is called mean-reverting. Two examples of the generalized Ornstein-Uhlenbeck process are considered. In the first one, we study the Ornstein-Uhlenbeck process on a comb model, as an example of the harmonically bounded random motion in the topologically constrained geometry. The main dynamical characteristics (as the first and the second moments) and the probability density function are studied in the framework of both the Langevin stochastic equation and the Fokker-Planck equation. The second example is devoted to the study of the effects of stochastic resetting on the Ornstein-Uhlenbeck process, including stochastic resetting in the comb geometry. Here, the non-equilibrium stationary state is the main question in task, where the two divergent forces, namely the resetting and the drift towards the mean, lead to compelling results both in the case of the Ornstein-Uhlenbeck process with resetting and its generalization on the two dimensional comb structure. © 2023, CC BY.

关键词： Brownian movement

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：