检索结果-内蒙古大学图书馆

Ivannikov ISPRAS Open Conference (ISPRAS)

作者： Rewer Canosa Andrei Tchernykh Jorge M. Cortés-Mendoza Raúl Rivera-Rodriguez Jose Lozano Rizk Arutyun Avetisyan Zhihui Du Gleb Radchenko Eduardo R. Concepción Morales CICESE Research Center Ensenada Mexico Institute for System Programming of the Russian Academy of Sciences Moscow Russia Tsinghua University Beijing China South Ural State University Chelyabinsk Russia Metropolitan University Quito Ecuador

ISBN: (纸本)9781728112763;9781728112756

Lightweight virtualization technology has emerged as an alternative to traditional hypervisor-based virtualization. Containers based on an operating system level virtualization have shown superior performance and more flexibility than virtual machines. Both factors encourage their fast adoption and wide use in cloud environments. Container technology guarantees efficient interaction with hardware through system calls and isolation of tasks execution. The portability and ease of deployment enable containers to execute a variety of jobs. Several techniques have been used to address the problem of job allocation on containers. In this paper, a set of allocation strategies is proposed to distribute the workload to multiple OS containers that are running on several servers. We address two objective scheduling problem: minimizing energy and a number of SLA violations. We propose seventy-two strategies and analyze their performance by comprehensive simulation. We demonstrate that proposed techniques outperform previous works in terms of both the energy savings and QoS.

关键词： 1/f noise Two dimensional displays

来源：评论

学校读者我要写书评

暂无评论

Evaluation of interpretability methods for adversarial robustness on real-world datasets

Evaluation of interpretability methods for adversarial robus...

引用

Ivannikov Memorial Workshop (IVMEM)

作者： Anna Chistyakova Maria Cherepnina Konstantin Arkhipenko Sergey D. Kuznetsov Chang-Seok Oh Sebeom Park Lomonosov Moscow State University Moscow Russia Ivannikov Institute for System Programming of the RAS Moscow Russia Moscow Institute of Physics and Technology (State University) Moscow Russia Plekhanov Russian University of Economics Moscow Russia Samsung Research Seoul South Korea

ISBN: (纸本)9781665423281

Adversarial training is considered the most powerful approach for robustness against attacks on deep neural networks involving adversarial examples. However, recent works have shown that the similar robustness level can be achieved by other means, namely interpretability-based regularization. We evaluate these interpretability-based approaches on real-world ResNet models trained on CIFAR-10 and ImageNet datasets. Our results show that interpretability can marginally improve robustness when combined with adversarial training, however, they bring additional computational complexity making these approaches questionable for such models and datasets.

关键词： Training Deep learning Computational modeling Conferences Neural networks Robustness Computational complexity

来源：评论

学校读者我要写书评

暂无评论

Local Methods with Adaptivity via Scaling

arXiv

引用

arXiv 2024年

作者： Chezhegov, Savelii Skorik, Sergey Khachaturov, Nikolas Shalagin, Danil Avetisyan, Aram Takáč, Martin Kholodov, Yaroslav Beznosikov, Aleksandr Moscow Institute of Physics and Technology Dolgoprudny Russia Ivannikov Institute for System Programming Moscow Russia Mohamed bin Zayed University of Artificial Intelligence Abu Dhabi United Arab Emirates Russian-Armenian University Yerevan Armenia Innopolis University Innopolis Russia

The rapid development of machine learning and deep learning has introduced increasingly complex optimization challenges that must be addressed. Indeed, training modern, advanced models has become difficult to implement without leveraging multiple computing nodes in a distributed environment. Distributed optimization is also fundamental to emerging fields such as federated learning. Specifically, there is a need to organize the training process to minimize the time lost due to communication. A widely used and extensively researched technique to mitigate the communication bottleneck involves performing local training before communication. This approach is the focus of our paper. Concurrently, adaptive methods that incorporate scaling, notably led by Adam, have gained significant popularity in recent years. Therefore, this paper aims to merge the local training technique with the adaptive approach to develop efficient distributed learning methods. We consider the classical Local SGD method and enhance it with a scaling feature. A crucial aspect is that the scaling is described generically, allowing us to analyze various approaches, including Adam, RMSProp, and OASIS, in a unified manner. In addition to the theoretical analysis, we validate the performance of our methods in practice by training a neural network. Copyright © 2024, The Authors. All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Verifying SDN data path requests

arXiv

引用

arXiv 2019年

作者： Burdonov, Igor Kossachev, Alexandre Yevtushenko, Nina López, Jorge Kushik, Natalia Zeghlache, Djamal Ivannikov Institute for System Programming of the Russian Academy of Sciences Moscow Russia National Research University Higher School of Economics Moscow Russia SAMOVAR CNRS Télécom SudParis Université Paris-Saclay Evry France

Software Defined Networking (SDN) is a pillar technology for network virtualization, which currently attracts a lot of attention due to the provided capabilities. In recent years, different works have been devoted to testing / verifying the (correct) configurations of SDN data planes. In general, SDN forwarding devices (e.g., switches) route (stir) traffic according to the configured flow rules;essentially, a flow rule decides which action to take (e.g., forward the received network packet to a set of ports) if the received network packet matches some predefined values. In this paper, we showcase misconfigurations which can occur due to the inherent working principles of flow rules. Namely, we discuss how, when synthesizing a set of data-paths, other data paths (including loops) may be unintentionally configured. Furthermore, we show that for some cases the original set of data paths cannot be implemented (only a superset of it). Additionally, we present a method for detecting such issues and estimate its complexity. The obtained results may be interesting for practical use due to their impact and low (polynomial) time complexity. Copyright © 2019, The Authors. All rights reserved.

关键词： Software defined networking

来源：评论

学校读者我要写书评

暂无评论

Speaker's Emotions Recognition Module Based on the GoogleLeNet Neural Network

Speaker's Emotions Recognition Module Based on the GoogleLeN...

引用

2022 International Conference on Smart Information systems and Technologies, SIST 2022

作者： Tereikovskyi, Ihor Tereikovska, Liudmyla Kryvoruchko, Olena Tereikovskyi, Oleh Tyshchenko, Dmytro Franchuk, Tamara National Technical University Of Ukraine Igor Sikorsky Kyiv Polytechnic Institute Dept. Of System Programming And Specialized Computer Systems Kyiv Ukraine Kyiv National University Of Construction And Architecture Dept. Of Information Technology For Design And Applied Mathematics Kyiv Ukraine Kyiv National University Of Trade And Economics Dept. Of Software Engineering And Cyber Security Kyiv Ukraine National Technical University Of Ukraine Igor Sikorsky Kyiv Polytechnic Institute Dept. Of Department Of Computer Engineering Kyiv Ukraine

ISBN: (数字)9781665467902

ISBN: (纸本)9781665467902

The article is devoted to the development of means for recognition of the emotions of the speaker, based on the neural network analysis of fixed fragments of the voice signal. The possibility of improving recognition tools through the use of a convolutional neural network of the GoogLeNet type is shown. The expediency of training the network on the examples of the TESS database is determined. A procedure for representing a voice signal in the form of a square grayscale image has been developed. Based on the use of the specified database, the modification of the GoogLeNet network was carried out. The architecture and software implementation of the recognition module have been developed. With the help of computer experiments, it has been established that after 150 epochs of training on a fairly limited training sample, the modified GoogLeNet makes it possible to achieve an accuracy of recognition of the speaker's emotions of approximately 0.97 on test examples. The obtained accuracy is comparable to the accuracy of the best modern systems of this purpose and confirms the promise of using neural network models like GoogLeNet for emotion recognition based on voice signal analysis. The need for further research aimed at reducing the resource intensity of the neural network model is shown. © 2022 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

The first optimal algorithm for smooth and strongly-convex-strongly-concave minimax optimization 22

The first optimal algorithm for smooth and strongly-convex-s...

引用

Proceedings of the 36th International Conference on Neural Information Processing systems

作者： Dmitry Kovalev Alexander Gasnikov King Abdullah University of Science and Technology Thuwal Saudi Arabia Moscow Institute of Physics and Technology Dolgoprudny Russia and Institute for System Programming RAS Research Center for Trusted Artificial Intelligence Moscow Russia and National Research University Higher School of Economics Moscow Russia

ISBN: (纸本)9781713871088

In this paper, we revisit the smooth and strongly-convex-strongly-concave minimax optimization problem. Zhang et al. (2021) and Ibrahim et al. (2020) established the lower bound $\Omega\left(\sqrt{\kappa_x\kappa_y} \log \frac{1}{\epsilon}\right)$ on the number of gradient evaluations required to find an ε-accurate solution, where kx and ky are condition numbers for the strong convexity and strong concavity assumptions. However, the existing state-of-the-art methods do not match this lower bound: algorithms of Lin et al. (2020) and Wang and Li (2020) have gradient evaluation complexity $\mathcal{O}\left(\sqrt{\kappa_x\kappa_y} \log^3 \frac{1}{\epsilon}\right)$ and $\mathcal{O}\left( \sqrt{\kappa_x\kappa_y}\log^3 (\kappa_x\kappa_y)\log\frac{1}{\epsilon}\right)$, respectively. We fix this fundamental issue by providing the first algorithm with O (√kxky log 1/ε) gradient evaluation complexity. We design our algorithm in three steps: (i) we reformulate the original problem as a minimization problem via the pointwise conjugate function; (ii) we apply a specific variant of the proximal point algorithm to the reformulated problem; (iii) we compute the proximal operator inexactly using the optimal algorithm for operator norm reduction in monotone inclusions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Intelligent system for Classifying of Acoustic Endoscopic Images of the Pancreas Based the Analysis of Walsh Spectra of Local Windows 25

Intelligent System for Classifying of Acoustic Endoscopic Im...

引用

25th International Arab Conference on Information Technology, ACIT 2024

作者： Al-Kasasbeh, Riad Taha Filist, Sergey Shatalova, Olga Kondrashov, Dmitry Sergeevich Shaqadan, Ashraf Korenevskiy, Nikolay A. Al-Habahbeh, Osama M. Tarawneh, Emad Saleh Starkov, Evgeny The University of Jordan Mechatronics Engineering Department Amman Jordan Moscow Polytechnic University Department of Smart Technologies Moscow Russia Southwest State University Department of Biomedical Engineering Kursk Russia Zarqa University Civil Engineering Department Zarqa Jordan University of Business and Technology Jeddah21448 Saudi Arabia South-West State University Russia The University of Jordan Jordan Institute of System Programming The Russian Academy of Sciences Russia

ISBN: (纸本)9798331540012

The purpose of the work is to develop and study methods and algorithms for intelligent classification systems for digital medical images. A method for automatic classification of medical images was developed using neural networks and transparency masks obtained by preliminary image segmentation. This allows you to generate descriptors only for those pixels that are selected by the transparency mask. Also algorithmic and software for classifying medical images has been developed. Experimental testing was carried out on ultrasound images of the pancreas. Differential diagnostics was carried out on control samples. The classes "oncology"and "pancreatitis"were divided. The diagnostic sensitivity by class turned out to be at least 0.8, while the diagnostic specificity is at least 0.85. © 2024 IEEE.

关键词： Ultrasonic imaging

来源：评论

学校读者我要写书评

暂无评论

Accelerated primal-dual gradient method for smooth and convex-concave saddle-point problems with bilinear coupling 22

Accelerated primal-dual gradient method for smooth and conve...

引用

Proceedings of the 36th International Conference on Neural Information Processing systems

作者： Dmitry Kovalev Alexander Gasnikov Peter Richtárik King Abdullah University of Science and Technology Thuwal Saudi Arabia Moscow Institute of Physics and Technology Dolgoprudny Russia and Institute for System Programming RAS Research Center for Trusted Artificial Intelligence Moscow Russia and National Research University Higher School of Economics Moscow Russia

ISBN: (纸本)9781713871088

In this paper we study the convex-concave saddle-point problem minx maxy f(x) + yTAx – g(y), where f(x) and g(y) are smooth and convex functions. We propose an Accelerated Primal-Dual Gradient Method (APDG) for solving this problem, achieving (i) an optimal linear convergence rate in the strongly-convex-strongly-concave regime, matching the lower complexity bound (Zhang et al., 2021), and (ii) an accelerated linear convergence rate in the case when only one of the functions f(x) and g(y) is strongly convex or even none of them are. Finally, we obtain a linearly convergent algorithm for the general smooth and convex-concave saddle point problem minx maxy F(x, y) without the requirement of strong convexity or strong concavity.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Optimizing Machine Learning Algorithms to Enhance Blood Donation for Liver Patients: A Case Study 7

Optimizing Machine Learning Algorithms to Enhance Blood Dona...

引用

7th IET Smart Cities Symposium, SCS 2023

作者： Mijwil, Maad M. El-Kenawy, El-Sayed M. Bala, Indu Aljanabi, Mohammad Abotaleb, Mostafa Ibrahim, Abdelhameed Computer Techniques Engineering Department Baghdad College of Economic Sciences University Baghdad Iraq Department of Communications and Electronics Delta Higher Institute of Engineering and Technology Egypt School of Electrical and Electronics Engineering Lovely Professional University Punjab India Department of Computer Aliraqia University Baghdad Iraq Department of System Programming South Ural State University Chelyabinsk Russia Computer Engineering and Control Systems Department Faculty of Engineering Mansoura University Mansoura Egypt

ISBN: (纸本)9781839539831

Blood is vital for transporting oxygen, nutrients, and hormones to all body parts as it circulates through arteries and veins. It removes carbon dioxide, regulates body temperature, and maintains the body's immune system. Individuals constantly need blood and its derivatives to save their lives and improve their health through medical treatments and surgical operations. Liver diseases are one of the diseases that affects the health of individuals and requires blood to continue living. These diseases cause significant damage to people's health, and early diagnosis plays a crucial role in saving lives. In this paper, machine learning algorithms (support vector machine and random forest) are involved in detecting liver diseases and determining whether donors are suitable to donate blood from blood values. This paper is applied research that found that the performance measures of the random forest algorithm achieved excellent performance in identifying suitable people to donate blood. © The Institution of Engineering & Technology 2023.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Comparative analysis of the scalar point multiplication algorithms in the NIST FIPS 186 elliptic curve cryptography 3

Comparative analysis of the scalar point multiplication algo...

引用

3rd International Workshop on Information, Computation, and Control systems for Distributed Environments, ICCS-DE 2021

作者： Babenko, Mikhail Tchernykh, Andrei Redvanov, Aziz Djurabaev, Anvar North-Caucasus Center for Mathematical Research North-Caucasus Federal University 1 Pushkin Street Stavropol355017 Russia Institute for System Programming of the Russian Academy of Sciences Moscow109004 Russia CICESE Research Center carr. Tijuana-Ensenada 3918 Ensenada22860 Mexico South Ural State University Prospekt Lenina 76 Chelyabinsk454080 Russia ITMO University 9 Lomonosova Street Saints-Petersburg191002 Russia

In today's world, the problem of information security is becoming critical. One of the most common cryptographic approaches is the elliptic curve cryptosystem. However, in elliptic curve arithmetic, the scalar point multiplication is the most expensive compared to the others. In this paper, we analyze the efficiency of the scalar multiplication on elliptic curves comparing Affine, Projective, Jacobian, Jacobi-Chudnovsky, and Modified Jacobian representations of an elliptic curve. For each coordinate system, we compare Fast exponentiation, Nonadjacent form (NAF), and Window methods. We show that the Window method is the best providing lower execution time on considered coordinate systems. Copyright © 2021 for this paper by its authors.

关键词： Security of data

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：