检索结果-内蒙古大学图书馆

Ivannikov ISPRAS Open Conference (ISPRAS)

作者： Maxim A. Ryndin Denis Y. Turdakov Sergey D. Kuznetsov Ivannikov Institute for System Programming of the RAS Moscow Russia Ivannikov Institute for System Programming of the RAS Lomonosov Moscow State University Moscow Russia Ivannikov Institute for System Programming of the RAS Lomonosov Moscow State University Moscow Institute of Physics and Technology (State University) National Research University Higher School of Economics Moscow Russia

ISBN: (数字)9781665412919

ISBN: (纸本)9781665448215

Modern supervised algorithms assume that the dataset used for training has the same distributions as the data to be processed. However, the real data is permanently changing. This leads to the gradual degradation of supervised machine learning algorithms in production systems and increases the cost of the maintaining. To solve this problem, we are focusing on domain adaptation of machine learning algorithms in lifelong manner. We assume that real unlabelled data come in continuously. For this setting we propose a method for detecting changes in data distributions, as well as updating supervised algorithms. The idea behind the method is to process a portion of the data and create a new labelled dataset for training a supervised model. The trained model becomes a part of the ensemble used for selecting a strategy to deal with new examples: assign the label automatically using co-training or manually with the aid of active learning. This method is independent of the specific architecture of the model and could be used with any modern supervised algorithms, including artificial neural networks. Our research also confirms two findings. First, adding small portion of data with reliable labels to a self-labelled dataset improves model's performance, even if this amount is small to build a model from scratch. It is also shown that accumulating domain knowledge by continuously adding new trained models to ensemble used for labelling, reduces the amount of labelled data required while maintaining the high performance of the adapted model.

关键词： Training Adaptation models Production systems Machine learning algorithms Data models Reliability Labeling

来源：评论

学校读者我要写书评

暂无评论

Neural network method for base extension in residue number system 2

Neural network method for base extension in residue number s...

引用

2nd International Workshop on Information, Computation, and Control systems for Distributed Environments, ICCS-DE 2020

作者： Babenko, M. Shiriaev, E. Tchernykh, A. Golimblevskaia, E. North-Caucasus Federal University Stavropol Russia CICESE Research Center Ensenada BC Mexico South Ural State University Chelyabinsk Russia Ivannikov Institute for System Programming Moscow Russia

Confidential data security is associated with the cryptographic primitives, asymmetric encryption, elliptic curve cryptography, homomorphic encryption, cryptographic pseudorandom sequence generators based on an elliptic curve, etc. For their efficient implementation is often used Residue Number system that allows executing additions and multiplications on parallel computing channels without bit carrying between channels. A critical operation in Residue Number system implementations of asymmetric cryptosystems is base extension. It refers to the computing a residue in the extended moduli without the application of the traditional Chinese Remainder Theorem algorithm. In this work, we propose a new way to perform base extensions using a Neural Network of a final ring. We show that it reduces 11.7% of the computational cost, compared with state-of-the-art approaches. Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

关键词： Computation theory

来源：评论

学校读者我要写书评

暂无评论

SoC approach for low cost and low power consumption based on ARM Cortex M3 according to ISO 26262

SoC approach for low cost and low power consumption based on...

引用

IEEE SSD International Multi-Conference on systems, Signals and Devices

作者： Muhammad Ikram Hafiz Mohamed Abdelawwad Waldemar Müller Eike Hahn Josef Börcsök ICAS Institute for Computer Architecture and System Programming University of Kassel Kassel Germany

ISBN: (纸本)9781665414937;9781665430579

The next milestone in the automotive industry is to achieve the complete autonomous driving capability. This demands high performance computing controllers and a larger number of Electronic Controller Units (ECUs) for realizing this goal. Also, the current shift towards electric cars where the battery capacity is limited, urges for new efficient approaches. Therefore, developing compact, miniaturized integrated controller that encompasses all the features is essential for minimizing the power consumption. At the same time, this approach of ECUs must also ensure the cost efficiency. To address and handle this need, the reduction in the ECUs numbers is necessary to reduce the cost factor but at the same time not compromising on the functionalities. Therefore, the approach of integrating these individual features in a single ECU is a promising way forward, i.e., system-on-Chip (SoC). To design and implement such an SoC for the automotive industry that ensures the safe operation of the controlled functions, it must follow a particular development criterion according to the ISO 26262 - functional safety of the road vehicles. This work focuses on the development of an SoC according to the ISO 26262 standard based on ARM Cortex M3 architecture which is an attractive candidate due to its low cost and low power consumption.

关键词： Industries Power demand ISO Standards Redundancy Layout Routing Safety

来源：评论

学校读者我要写书评

暂无评论

Multi-cloud Privacy-Preserving Logistic Regression 7th

Multi-cloud Privacy-Preserving Logistic Regression

引用

7th Russian Supercomputing Days Conference, RuSCDays 2021

作者： Cortés-Mendoza, Jorge M. Tchernykh, Andrei Babenko, Mikhail Pulido-Gaytán, Bernardo Radchenko, Gleb South Ural State University Chelyabinsk Russia CICESE Research Center BC Ensenada Mexico Ivannikov Institute for System Programming Moscow Russia North-Caucasus Federal University Stavropol Russia

ISBN: (纸本)9783030928636

Clouds can significantly reduce the cost and time of business solutions. However, cloud services introduce significant security and privacy challenges when they process sensitive information. For instance, a dataset for machine learning could contain delicate information that traditional encryption approaches cannot protect during data analysis. Homomorphic Encryption (HE) schemes and secure Multi-Party Computation (MPC) are considered solutions for privacy protection in third-party infrastructures. In this paper, we propose a Multi-Cloud Logistic Regression based on Residue Number system (MC-LR-RNS) that provides security, parallel processing, and scalability. To validate the efficiency and practicability of the solution, we provide its analysis with different configurations, datasets, and cloud service providers. We use six available datasets from medicine (diabetes, cancer, drugs, etc.) and genomics. The analysis shows that MC-LR-RNS provides the same levels of quality as non-HE solutions and improved performance due to multi-cloud parallel computations. © 2021, Springer Nature Switzerland AG.

关键词： Distributed database systems

来源：评论

学校读者我要写书评

暂无评论

Numerical simulation of irregular waves by HOS method 32nd

Numerical simulation of irregular waves by HOS method

引用

32nd International Ocean and Polar Engineering Conference, ISOPE 2022

作者： Huang, Congyi Xu, Ping Wan, Decheng Strijhak, Sergei School of Naval Architecture Ocean and Civil Engineering Shanghai Jiao Tong University Shanghai China Marine Design and Research Institute of China Shanghai China Ivannikov Institute for System Programming of the Russian Academy of Sciences Moscow Russia

ISBN: (纸本)9781880653814

Ships and offshore structures often encounter irregular waves in the ocean. Therefore, the simulation of irregular waves is very necessary and meaningful. The potential flow method is often used to simulate the irregular waves because the evolution needs a long time and a large amount of calculation. High order spectral (HOS) method is a kind of potential flow method, which has the advantages of high efficiency. In this paper, the HOS method is used to simulate the formation and evolution of irregular waves. Firstly, the governing equations, boundary conditions and the solving procedure of the HOS-ocean model based on the HOS method are introduced. Then the accuracy and stability of the HOS-ocean model are verified by a standard example. After that, the formation and evolution of two-dimensional and three-dimensional irregular waves are simulated based on JONSWAP spectrum and ITTC spectrum respectively. The frequency domain curves, time domain curves, maximum wave height, maximum wave steepness, average wave steepness and other parameters of the waves generated by the two wave spectrums are compared, and the differences between the two wave spectrums in generating irregular waves are analyzed. © 2022 by the International Society of Offshore and Polar Engineers (ISOPE).

关键词： Potential flow

来源：评论

学校读者我要写书评

暂无评论

Survey of methods for automated code-reuse exploit generation

arXiv

引用

arXiv 2020年

作者： Vishnyakov, Alexey V. Nurmukhametova, Alexey R. Ivannikov Institute for System Programming Russian Academy of Sciences Moscow109004 Russia

This paper provides a survey of methods and tools for automated code-reuse exploit generation. Such exploits use code that is already contained in a vulnerable program. The code-reuse approach allows one to exploit vulnerabilities in the presence of operating system protection that prohibits data memory execution. This paper contains a description of various code-reuse methods: return-to-libc attack, return-oriented programming, jump-oriented programming, and others. We define fundamental terms: gadget, gadget frame, gadget catalog. Moreover, we show that, in fact, a gadget is an instruction, and a set of gadgets defines a virtual machine. We can reduce an exploit creation problem to code generation for this virtual machine. Each particular executable file defines a virtual machine instruction set. We provide a survey of methods for gadgets searching and determining their semantics (creating a gadget catalog). These methods allow one to get the virtual machine instruction set. If a set of gadgets is Turing-complete, then a compiler can use a gadget catalog as a target architecture. However, some instructions can be absent. Hence we discuss several approaches to replace missing instructions with multiple gadgets. An exploit generation tool can chain gadgets by pattern searching (regular expressions) or considering gadgets semantics. Furthermore, some chaining methods use genetic algorithms, while others use SMT-solvers. We compare existing open-source tools and propose a testing system rop-benchmark that can be used to verify whether a generated chain successfully opens a shell. © 2020, CC BY-NC-ND.

关键词： Surveys

来源：评论

学校读者我要写书评

暂无评论

Towards Optimizing Cloud Computing Using Residue Number system

Towards Optimizing Cloud Computing Using Residue Number Syst...

引用

International Conference on Marchuk Scientific Readings 2020, MSR 2020

作者： Kucherov, N. Kuchukova, E. Tchernykh, A. Kuchukov, V. Babenko, M. North-Caucasus Federal University Stavropol Russia CICESE Research Center Ensenada BC Mexico South Ural State University Chelyabinsk Russia Ivannikov Institute for System Programming Moscow Russia

In the work methods for building systems of distributed data storage based on the system of residual classes are considered. The use of direct conversion of data from the positional system of calculation to the residue number system will have a large computational complexity, the use of modules of a particular type allows you to solve this problem. The operation of scaling and expansion of the base system, which is necessary to restore the number of stored parts in case of failure of one or more cloud servers, is considered. © 2021 institute of Physics Publishing. All rights reserved.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

Communication SoC based on 1oo2D architecture for industrial human-robot-collaboration

Communication SoC based on 1oo2D architecture for industrial...

引用

IEEE SSD International Multi-Conference on systems, Signals and Devices

作者： Mohamed Abdelawwad Malte Drabesch Michael Schwarz Muhammad Ikram Hafiz Josef Börcsök ICAS Institute for Computer Architecture and System Programming University of Kassel Kassel Germany

ISBN: (纸本)9781665414937;9781665430579

The mechanical force imparted by industrial robots is enormous and can lead to dangerous situations for human workers and property if they malfunction. system designs based on the redundancy concept are essential to avoid or mitigate the resulting risk, which is recommended by the various safety standards. In this paper, a 1oo2D based safety SoC that provides multiple Ethernet communication interfaces is presented. Two different design approaches are implemented using the in-house FPGA development board based on the LEON3 soft processor. The software algorithms for each approach are developed and verified.

关键词： Service robots Software algorithms Ethernet Computer architecture Robot sensing systems Software Safety

来源：评论

学校读者我要写书评

暂无评论

High-resolution simulation of internal waves attractors and impact of interaction of high amplitude internal waves with walls on dynamics of waves attractors 7

High-resolution simulation of internal waves attractors and ...

引用

7th European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS Congress 2016

作者： Sibgatullin, Ilias Kalugin, Michael Institute of System Programming Solzhenitsina str. 25 Moscow Russia Moscow State University Faculty of Mechanics and Mathematics Main Building GSP-1 1 Leninskiye Gory Moscow119991 Russia

ISBN: (纸本)9786188284401

Internal wave attractors have received great attention since its discovery in 1995 by Leo Maas ([1, 2]). Now convectional theory describing the formation of attractors is generally accepted, and the principal interest of researchers in recent years is focused on nonlinear interactions. A number of theories had been proposed and now nonlinear interaction due to triadic resonance is accepted as a principle cause of instability of attractors, in which case the parent wave of large amplitude gives birth to two daughter waves, such that conditions of triadic resonance are fulfilled for all the three waves [3]. All this presumes that wave-wall interaction participate in the process only by focusing-defocusing of energy on the inclined boundaries. Here we address interactions of large amplitude internal waves with the boundaries in real laboratory conditions, where Prandtl-Schmidt number is equal to 700. We show that large amplitude waves produce folded structures which are clearly visible on density-gradient images, as if produced by "kneading the dough". These structures are not quickly dissipated due to high Schmidt number and with time they propagate to the interior of the domain, as was shown by our numerical simulation. These structures have visible impact on the instability of attractor and the whole picture of turbulent motion for large amplitudes. The mechanism of interaction of these structures with internal gravity waves structures is the subject of further research. Numerical simulation is quite challenging due to high Prandtl-Schmidt number and small scale of the folded structures in highly nonlinear regimes. Also study is performed on large time scales. As a consequence most of conventional computational approaches give unreliable results. We have applied spectral element approach base on code nek5000 by Paul Fischer. It allowed us to carefully follow the fine space structures on large time scales.

关键词： Numerical models

来源：评论

学校读者我要写书评

暂无评论

Random Graph Modeling: A survey of the concepts

arXiv

引用

arXiv 2024年

作者： Drobyshevskiy, Mikhail Turdakov, Denis Ivannikov Institute for System Programming The Russian Academy of Sciences Moscow Institute of Physics and Technology Russia

Random graph (RG) models play a central role in the complex networks analysis. They help to understand, control, and predict phenomena occurring, for instance, in social networks, biological networks, the Internet, etc. Despite a large number of RG models presented in the literature, there are few concepts underlying them. Instead of trying to classify a wide variety of very dispersed models, we capture and describe concepts they exploit considering preferential attachment, copying principle, hyperbolic geometry, recursively defined structure, edge switching, Monte Carlo sampling, etc. We analyze RG models, extract their basic principles, and build a taxonomy of concepts they are based on. We also discuss how these concepts are combined in RG models and how they work in typical applications like benchmarks, null models, and data anonymization. © 2024, CC BY-NC-ND.

关键词： Benchmarking

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：