检索结果-内蒙古大学图书馆

39th International Conference on High Performance Computing, ISC High Performance 2024

作者： Ratsaby, Joel Timashkov, Alexander Electrical and Electronics Engineering Department Ariel University Ariel Israel

ISBN: (纸本)9783982633602

We introduce a method for processing unstructured data for machine learning based on an LZ-complexity string distance. Computing the LZ-complexity is inherently a serial data compression process;hence, we introduce a string distance computed by a parallel algorithm that utilizes multiple GPU devices to process unstructured data, which typically exists in large quantities. We use this algorithm to compute a distance matrix representation of the unstructured data that standard learning algorithms can use to learn. Our approach eliminates the need for human-based feature definition or extraction. Except for some simple data reformatting done manually, our proposed approach operates on the original raw data and is fully automatic. The parallel computation of the distance matrix is efficient. It obtains a speed-up factor of 528 in computing the distance matrix between every possible pair of 16 strings of length 1M bytes. We show that for learning time-series classification, relative to the ubiquitous TFIDF data representation, the distance-matrix representation yields a higher learning accuracy for most of a broad set of learning algorithms. Thus, the parallel algorithm can be helpful in efficiently and accurately learning from unstructured data. © 2024 Research Paper Proceedings of the ISC High Performance 2024. All rights reserved.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

An Efficient Algorithm for Continuous Complex Event Matching Using Bit-parallelism 40

An Efficient Algorithm for Continuous Complex Event Matching...

引用

40th IEEE International Conference on Data Engineering, ICDE 2024

作者： Qiu, Tao Jiang, Shenwang Yang, Xiaochun Wang, Bin Zong, Chuanyu Zhu, Rui Shenyang Aerospace University China Northeastern University China

ISBN: (纸本)9798350317152

Complex event matching has gained a lot of at-tention for evaluating complex queries over event streams. The events composing a complex event occur within a user-specified time window and can be nonconsecutive on the stream. Existing methods widely utilize the state automaton to match complex events. However, the state automaton is typically used for matching consecutive items satisfying a pattern, e.g., the regular expression. To support nonconsecutive event matching, it has to maintain a large number of partial matches and skip irrelevant events, which results in a huge overhead. To avoid this problem, we employ the bit parallelism technique to match complex events continuously in this paper. We utilize a set of bit sequences to represent the events, where each bit is associated with a time slice, and an event is mapped to a 1-bit of the sequence if its timestamp belongs to the time slice. Then, bit-parallel operations are designed to process the constraints defined on the complex event, e.g., the time window limitation, and sequential order of the events, etc. We further propose the bit-parallel algorithms to support continuous complex event matching using these bit operations. Our experiments on real and synthetic datasets demonstrate that our method outperforms the existing methods by up to an order of magnitude in Query efficiency. © 2024 IEEE.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

parallel Segmentation Algorithm Based on Computation Front for Numerical Acceleration in Electrical Drive Real-time Simulation 5

Parallel Segmentation Algorithm Based on Computation Front f...

引用

5th IEEE Student Conference on Electric Machines and Systems, SCEMS 2022

作者： Zhang, Zhe He, Shaomin Chen, Jian Yang, Huan Zhao, Rongxiang Yang, Ruoyan Zhejiang University College of Electrical Engineering Hangzhou China Shenzhen Costoms District Shenzhen China

ISBN: (数字)9781665476898

ISBN: (纸本)9781665476898

HIL (Hardware in Loop) is an efficient and convenient tool for the test and verification of electrical drive system which requires high reliability and safety. With the application of high-frequency SiC inverter, the time scale is reduced below μs-level, as a result, the computation speed becomes an unignorable challenge for HIL system because of the strict real-time constraint. Under these circumstances, the development and applications of hardware with parallel process structure such as FPGA is a solution for computation acceleration on the hardware level. On the software level, unfortunately, most of the state-of-the-art numerical models and algorithms are based on sequential mechanism, as a result, the parallel algorithm is demanded to achieve further acceleration. This paper proposes a general method based on computation front to design parallel algorithm for real-time simulation. By analyzing numerical flow diagram, numerical integration method is optimized to realize acceleration derived from parallel segmentation. Simulation results show that good acceleration effect without great accuracy reduction is achieved. © 2022 IEEE.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

ML-based Approach for Credit Risk Assessment Using parallel Calculations 3

ML-based Approach for Credit Risk Assessment Using Parallel ...

引用

3rd International Workshop on Computational and Information Technologies for Risk-Informed Systems, CITRisk 2022

作者： Hentosh, Lesia Tsikalo, Yevhen Kustra, Natalya Kutucu, Hakan Department of Artificial intelligence Lviv Polytechnic National University Lviv79013 Ukraine Department of Accounting and Audit Ivan Franko National University of Lviv Lviv79008 Ukraine Department of Publishing Information Technologies Lviv Polytechnic National University Lviv79013 Ukraine Department of Software Engineering Karabuk University Karabuk78050 Turkey

In banks and other credit organizations, the task of credit scoring often arises when making decisions on granting loans. The last one consists of making a reasoned decision based on information about the applicant, whether she should be granted a loan, and, if so, on what terms. This paper proposes the application of parallel calculations of the Random forest algorithm when solving the credit scoring task. This approach made it possible to reduce the time of model training and dataset processing significantly. Expectedly, when applying less data, the resulting acceleration and efficiency worsen. Using only 2500 entries, the execution time of the sequential algorithm is less than the parallel algorithm. The developed software was tested on three different processors: 4-core, 8-core, and 12-core, to evaluate the parallelization quality of data pre-processing. The classification algorithm is computationally complex and time-consuming, so we obtained practically the same acceleration for processing 5000 and 10000 records. With this amount of data, the 12-core processor gave the biggest gain in time when working with 12 threads. As a result, it is possible to have an acceleration of more than 6. This efficiency indicator of the proposed parallel algorithm can be significantly improved by varying the number of threads and considering the current trends in developing the multi-core architecture of computing systems. Also, using data without pre-processing, the following evaluation metrics were obtained: AUC=0.9 and Precision=0.845, and using data after pre-processing, these metrics were: AUC=0.86, Precision=0.89. © 2022 Copyright for this paper by its authors.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

A Hybrid parallel Algorithm With Multiple Improved Strategies 12th

A Hybrid Parallel Algorithm With Multiple Improved Strategie...

引用

12th IFIP TC 12 International Conference on Intelligent Information Processing, IIP 2022

作者： Wang, Tingting Pan, Jeng-Shyang Song, Pei-Cheng Chu, Shu-Chuan College of Computer Science and Engineering Shandong University of Science and Technology Qingdao China Department of Information Management Chaoyang University of Technology Taichung Taiwan College of Science and Engineering Flinders University Bedford Park Australia

ISBN: (纸本)9783031039478

This paper proposes a novel hybrid parallel algorithm with multiple improved strategies. The whole population is divided into three subpopulations and each sub-population executes butterfly optimization algorithm, grey wolf optimization algorithm, and marine predator algorithm respectively. Meanwhile, they share information through three different communication strategies. And in order to improve the performance of the algorithm, the text uses the cubic chaotic mapping mechanism in the initialization stage. At the same time, the idea of adaptive parameter strategy is also introduced, so that some hyperparameters are changed along with the iteration. The results show that the algorithm can provide very competitive results, and is superior to the best algorithm in the literature on most test functions. © 2022, IFIP International Federation for Information Processing.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

Research on parallel Algorithm of High-Power Microwave Device Simulation Based on MSMPI 23

Research on Parallel Algorithm of High-Power Microwave Devic...

引用

23rd International Vacuum Electronics Conference, IVEC 2022

作者： Hu, Yulan Liu, Dagang Liu, Laqun Wang, Huihui School of Electronic Science and Engineering University of Electronic Science and Technology of China Chengdu610054 China

ISBN: (纸本)9781665443258

The improvement of parallel computing efficiency is of great significance to the development and research of particle simulation theory. This article first compares the communication efficiency of MSMPI and MPICH2. The electromagnetic particle-in-cell simulation parallel algorithm and dynamic load balancing algorithm based on MSMPI shared memory characteristics are designed. Finally, the parallel test of RKA and MILO was realized by the particle simulation software CHIPIC-3D. The test results show that the two algorithms can significantly improve the efficiency of parallel computing. © 2022 IEEE.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

Improving the efficiency of sparse matrix class processing by using the SPM-CSR parallel algorithm and OpenMP technology 4

Improving the efficiency of sparse matrix class processing b...

引用

4th International Youth Conference on Radio Electronics, Electrical and Power Engineering, REEPE 2022

作者： Al-Saeedi, Adnan Adhab K. Yurievna, Shamaeva Olga Khairi, Teaba W. National Research University 'MPEI' Department of Applied Mathematics and Artificial Intelligence Moscow Russia Al-Furat AL-Awast Technical University Babylon Technical Institute Kufa Iraq University of Technology-Iraq Computer Science Dept Iraq

ISBN: (纸本)9781665414340

The development of computationally efficient algorithms and the improvement of their software implementation are urgent issues that require continuous attention due to the ongoing development of computer system architecture. In multithreading systems bad coordination cause a big problems and give unwilling results. The purpose of the work is to improve performance of the parallel algorithm for sparse matrix multiplication in CSR format by using vector (SpMV-CSR) using OpenMP technology. Using OpenMP technology and reduction directive to solve sparse matrix time execution and bad coordination problem leads to execution time reduction and speed-up increase. Results obtained an acceleration increase of 14% when using reduction. © 2022 IEEE.

关键词： parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

Work-Stealing Prefix Scan: Addressing Load Imbalance in Large-Scale Image Registration

引用

IEEE TRANSACTIONS ON parallel AND DISTRIBUTED SYSTEMS 2022年第3期33卷 523-535页

作者： Copik, Marcin Grosser, Tobias Hoefler, Torsten Bientinesi, Paolo Berkels, Benjamin Swiss Fed Inst Technol Dept Comp Sci CH-8092 Zurich Switzerland Univ Edinburgh Sch Informat Edinburgh EH8 9YL Midlothian Scotland Umea Univ Dept Comp Sci S-90736 Umea Sweden Rhein Westfal TH Aachen Inst Geometry & Pract Math AICES Grad Sch D-52062 Aachen Germany

parallelism patterns (e.g., map or reduce) have proven to be effective tools for parallelizing high-performance applications. In this article, we study the recursive registration of a series of electron microscopy images - a time consuming and imbalanced computation necessary for nano-scale microscopy analysis. We show that by translating the image registration into a specific instance of the prefix scan, we can convert this seemingly sequential problem into a parallel computation that scales to over thousand of cores. We analyze a variety of scan algorithms that behave similarly for common low-compute operators and propose a novel work-stealing procedure for a hierarchical prefix scan. Our evaluation shows that by identifying a suitable and well-optimized prefix scan algorithm, we reduce time-to-solution on a series of 4,096 images spanning ten seconds of microscopy acquisition from over 10 hours to less than 3 minutes (using 1024 Intel Haswell cores), enabling derivation of material properties at nanoscale for long microscopy image series.

关键词： Microscopy Image registration Heuristic algorithms Scanning electron microscopy Load management Standards parallel algorithms Prefix sum parallel algorithms work stealing load balancing image registration

来源：评论

学校读者我要写书评

暂无评论

Exploiting soft constraints within decomposition and coordination methods for sub-hourly unit commitment

引用

INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS 2022年 139卷 108023-108023页

作者： Raghunathan, Niranjan Bragin, Mikhail A. Yan, Bing Luh, Peter B. Moslehi, Khosrow Feng, Xiaoming Yu, Yaowen Yu, Chien-Ning Tsai, Chia-Chun Univ Connecticut Dept Elect & Comp Engn Storrs CT 06269 USA Rochester Inst Technol Dept Elect & Microelect Engn Rochester NY 14623 USA Hitachi ABB Power Grids San Jose CA 95134 USA Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automation Wuhan 430074 Peoples R China

Unit commitment (UC) is an important problem solved on a daily basis within a strict time limit. While hourly UC is currently used, they may not be flexible enough to accommodate the growing changes of demand and the increasing penetration of intermittent renewables. Sub-hourly UC is therefore recommended. This, however, will significantly increase problem complexity even under the deterministic setting because of the considerable increase of the number of intervals, leading to the drastic increase of the numbers of system coupling constraints and binary variables as compared to that of hourly UC. Consequently, existing methods may not be able to obtain good solutions within the time limit for large problems. In this paper, deterministic sub-hourly UC is considered with an innovative exploitation of "soft constraints" constraints that do not need to be strictly satisfied but their violations are penalized by predetermined coefficients. This, in conjunction with our recent "Surrogate Absolute Value Lagrangian Relaxation" approach where the "relaxed problem" is not required to be fully optimized, facilitates the formation and resolution of a new type of subproblems where soft system coupling constraints (e.g., reserve and transmission capacity constraints) are not relaxed. This then leads to a drastic reduction of the number of multipliers, decreased computational requirements, and improved solution quality. To further enhance the speed, a parallel version is developed. Testing results based on the Polish system demonstrate the effectiveness and robustness of both the sequential and parallel versions at finding high-quality solutions within the time limit.

关键词： Sub-hourly unit commitment Soft constraints Surrogate Lagrangian Relaxation (SLR) Surrogate Absolute-Value Lagrangian Relaxa-tion (SAVLR) parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

Optimization of a GMRES-based parallel algorithm for numerical simulation of multicomponent multiphase flow in porous media

Optimization of a GMRES-based parallel algorithm for numeric...

引用

European Conference on the Mathematics of Geological Reservoirs 2022, ECMOR 2022

作者： Kassymbek, N. Lebedev, D. Akhmed-Zaki, D. Kazakh National University Kazakhstan Astana IT University Kazakhstan

ISBN: (纸本)9789462824263

The actual task of petroleum geophysics is to solve the problem of multicomponent multiphase flow in a porous medium. At the same time, the development of effective parallel algorithms is an urgent task for modeling processes of this type. For these purposes, the authors carried out the following works. The problem of modeling the movement of a multicomponent multiphase liquid in a porous medium was solved. The equations describing fluid movement were linearized by the Newton-Raphson method and the resulting system of linear equations was solved using the generalized minimal residuals method (GMRES) with the Incomplete Lower-Upper factorization (ILU(0)) preconditioner. As a result, the problem was solved by the developed algorithm called Newton-ILU(0)-GMRES. A parallel program using the MPI standard and a fragmented program in the LuNA language for shared memory systems were developed and implemented. After analyzing the results of the previous work, it was decided to optimize the parallel GMRES algorithm, since the parallel program was not effective enough. Bottlenecks were identified, in particular Arnoldi orthogonalization, which negatively affected the effectiveness of the parallel program. An optimized version of the GMRES parallel algorithm was developed and implemented, which avoids the loss of parallel execution efficiency on Arnoldi orthogonalization. The developed program was tested and the results were analyzed. Comparisons were made with the previous version of the developed method. © 2022 European Conference on the Mathematics of Geological Reservoirs 2022, ECMOR 2022. All rights reserved.

关键词： parallel algorithms

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：