检索结果-内蒙古大学图书馆

19th Pacific Rim International Conference on Artificial Intelligence (PRICAI)

作者： Zhang, Rui Lensen, Andrew Sun, Yanan Sichuan Univ Chengdu 610000 Peoples R China Victoria Univ Wellington Wellington 6140 New Zealand

ISBN: (纸本)9783031208614;9783031208621

Symbolic regression has multiple applications in data mining and scientific computing. Genetic Programming (GP) is the mainstream method of solving symbolic regression problems, but its execution speed under large datasets has always been a bottleneck. This paper describes a CUDA-based parallel symbolic regression algorithm that leverages the parallelism of the GPU to speed up the fitness evaluation process in symbolic regression. We make the fitness evaluation step fully performed on the GPU and make use of various GPU hardware resources. We compare training time and regression accuracy between the proposed approach and existing symbolic regression frameworks including gplearn, TensorGP, and KarooGP. The proposed approach is the fastest among all the tested frameworks in both synthetic benchmarks and large-scale benchmarks.

关键词： Symbolic regression Genetic programming parallel algorithm Graphics processing unit (GPU) Compute unified device architecture (CUDA)

来源：评论

学校读者我要写书评

暂无评论

A New Link Prediction algorithm Based on Local Links

A New Link Prediction Algorithm Based on Local Links

引用

16th International Conference on Web-Age Information Management (WAIM)

作者： Yang, Juan Yang, Lixin Zhang, Pengye Beijing Univ Posts & Telecommun Key Lab Intelligent Telecommun Software & Multime Beijing 100876 Peoples R China

ISBN: (纸本)9783319235318;9783319235301

Link prediction refers to estimating the possibility of the existence of non-existent links between the nodes. The link prediction algorithms based on local information merely consider nodes' attributes or a small amount of topology information about common neighbors. In this paper, we proposed a new measure motivated by the cohesion between common neighbors and the predicted nodes-LNL (Local Neighbors Link). Experiments show that, compared with four classical algorithms on seven real networks, LNL has the higher accuracy and robustness. Furthermore, we apply the link prediction algorithms into large-scale networks. We implement the LNL method in both MapReduce and Spark, the experiments show that the implementation by Spark has higher efficiency than using MapReduce.

关键词： Link prediction Complex network Adjacent nodes parallel algorithm Spark

来源：评论

学校读者我要写书评

暂无评论

A single-tree algorithm to compute the Euclidean minimum spanning tree on GPUs 51

A single-tree algorithm to compute the Euclidean minimum spa...

引用

51st International Conference on parallel Processing (ICPP)

作者： Prokopenko, Andrey Sao, Piyush Lebrun-Grandie, Damien Oak Ridge Natl Lab Oak Ridge TN 37830 USA

ISBN: (纸本)9781450397339

Computing the Euclidean minimum spanning tree (Emst) is a computationally demanding step of many algorithms. While work-efficient serial and multithreaded algorithms for computing Emst are known, designing an efficient GPU algorithm is challenging due to a complex branching structure, data dependencies, and load imbalances. In this paper, we propose a single-tree Boruvka-based algorithm for computing Emst on GPUs. We use an efficient nearest neighbor algorithm and reduce the number of the required distance calculations by avoiding traversing subtrees with leaf nodes in the same component. The developed algorithms are implemented in a performance portable way using ArborX, an open-source geometric search library based on the Kokkos framework. We evaluate the proposed algorithm on various 2D and 3D datasets, show and compare it with the current state-of-the-art open-source CPU implementations. We demonstrate 4-24x speedup over the fastest multi-threaded implementation. We prove the portability of our implementation by providing results on a variety of hardware: AMD EPYC 7763, Nvidia A100 and AMD MI250X. We show scalability of the implementation, computing Emst for 37 million 3D cosmological dataset in under a 0.5 second on a single A100 Nvidia GPU.

关键词： Euclidean minimum spanning tree parallel algorithm GPU

来源：评论

学校读者我要写书评

暂无评论

parallel Computation of the 3-D Euclidean Distance Transform on the SIMD Hypercube Computer

Parallel Computation of the 3-D Euclidean Distance Transform...

引用

The Fourth International Conference on parallel and Distributed Computing, Applications and Technologies

作者： Yu-Hua Lee Shi-Jinn Horng Department of Computer Science and Information Engineering Ching Yun Instutute of Technology Department of Computer Science and Information Engineering National Taiwan University of Science and Technology

The processing of three-dimensional（3-D） objects from 3-D digital image data is an important task in the image processing and the computer vision *** distance transform（DT） is extensively applied in the image processing and computer vision areas as a key operation. In a two or three-dimensional image array,the computation of distance transform（DT） is an important task. With the increasing application of 3D voxel images,it is useful to consider the distance transform of a 3D digital image *** order to provide the efficient transform computations, parallelism is *** develop parallel algorithms for the three-dimensional Euclidean distance transform （3D-EDT） on the SIMD hypercube *** time complexity of our parallel algorithm is O（log N） for an N×JV×N image array using JV3 processors.A generalized parallel algorithm for the 3D-EDT is also proposed and it runs O(（N/p） log（N） +（N/p） log p) time for an N×N×N binary image array on the SIMD hypercube computer using p PE's,where 1≤p≤N.

关键词： Computer Vision Distance Transform Image Processing parallel algorithm SIMD Hypercube Computer

来源：评论

学校读者我要写书评

暂无评论

parallel Implementation of the Coordinates-partitioning Based Aggregation-type Algebraic Multigrid Preconditioners

Parallel Implementation of the Coordinates-partitioning Base...

引用

2017 International Conference on Mathematics, Modelling and Simulation Technologies and Applications（MMSTA 2017)

作者： Jian-ping WU Jun ZHAO Shu-chang WANG College of Meteorology and Oceanography National University of Defense Technology

The coordinates-partitioning based aggregation-type algebraic multigrid preconditioners have been proven to be very efficient in the solution of sparse linear systems with conjugate gradient *** this paper,a parallel algorithm for the setup is provided and the parallelization of the preconditioning process is also *** parallel algorithm is based on a good property of the coordinates-partitioning based aggregation,that is,the aggregation process is performed in a fashion from the coarsest level to the finest step by ***,the original adjacent graph is partitioned into a number of sub-graphs,where each sub-graph is related to a node in the coarsest level and is assigned to a *** aggregation process can then proceed forward on each processor independently from the assigned *** this kind of multigrid preconditioner is applied in Krylov subspace iterations,only the computation on the coarsest level and the matrix-vector multiplications related to the smoothing on each level require communication for V-and W-cycle versions,and require only some extra communication related to dot products for the K-cycle *** size of the derived linear system on the coarsest level can be controlled by the adjustable arguments and this system can be solved again in parallel with some preconditioned Krylov subspace *** structure information related to the matrix-vector multiplications is invariable to the iterations and can be derived in the setup and be stored,and is used in the latter ***,the parallel algorithm is validated for the Gauss-Seidel smoother and some kinds of popular multigrid cycles in solving sparse linear systems from some two-dimensional model partial equations with preconditioned conjugate gradient *** results show that the parallel efficiencies of both the setup and the iteration processes are satisfied.

关键词： Aggregation based algebraic multigrid Sparse linear system Coordinates partitioning Preconditioner Conjugate gradient method parallel algorithm

来源：评论

学校读者我要写书评

暂无评论

Primal-dual algorithms for the Capacitated Single Allocation p-Hub Location Problem

引用

International Journal of Hybrid Intelligent Systems 2022年第1-2期18卷 1-17页

作者： Matos, Telmo Ciicesi Escola Superior de Tecnologia e Gestão Politécnico Do Porto Porto Portugal

The Hub Location Problems (HLP) have gathered great interest due to the complexity and to the many applications in industry such as aviation, public transportation, telecommunications, among others. The HLP have many variants regarding allocation (single or multiple) and capacity (uncapacitated or capacitated). This paper presents a variant of the HLP, encompassing single allocation with capacity constraints. The Capacitated Single Allocation p-Hub Location Problem (CSApHLP) objective consists on determine the set of p hubs in a network that minimizes the total cost of allocating all the non-hub nodes to the p hubs. In this work, it is proposed a sophisticated RAMP approach (PD-RAMP) to improve the results obtained previously by the simple version (Dual-RAMP). Thus, a parallel implementation is conducted to assess the effectiveness of a parallel RAMP model applied to the CSApHLP. The first algorithm, the sequential PD-RAMP, incorporates Dual-RAMP with a Scatter Search procedure to create a Primal-Dual RAMP approach. The second algorithm, the parallel PD-RAMP, also take advantage of the dual and primal, parallelizing the primal side of the problem and interconnecting both sides as it is expected in the RAMP sequential algorithm. The quality of the results carried out on a standard testbed shows that the PD-RAMP approach managed to improve the state-of-the-art algorithms for the CSApHLP. © 2022 - IOS Press. All rights reserved.

关键词： CSApHLP lagrangean relaxation parallel algorithm RAMP

来源：评论

学校读者我要写书评

暂无评论

A divide & conquer approach to testing concurrent programs with JPF 27

A divide & conquer approach to testing concurrent programs w...

引用

27th Asia-Pacific Software Engineering Conference (APSEC)

作者： Do, Canh Minh Ogata, Kazuhiro Japan Adv Inst Sci & Technol Sch Informat Sci Nomi Ishikawa Japan

ISBN: (纸本)9781728195537

Java Pathfinder (JPF) is one of the most mature software model checker developed by NASA for detecting errors lurking in concurrent Java programs. However, the use of JPF often leads to the state space explosion due to the non-determinism of thread interleavings. parallelization is one mainstream technique to deal with the problem by utilizing the advantage of multi-core processors. Hence, making JPF parallelized to improve the performance in verifying programs is highly desired. In this paper, we present a divide & conquer approach to splitting the entire state space into multiple layers, in which each layer consists of many sub-state spaces so that JPF can independently verify such sub-state spaces in parallel. Some experiments demonstrate that the proposed technique mitigates the state space explosion that prevents the use of only one JPF instance from conducting model checking.

关键词： Concurrent programs Divide & conquer parallel algorithm Java Pathfinder (JPF)

来源：评论

学校读者我要写书评

暂无评论

Multi-modal search with convex bounding neighbourhood

Multi-modal search with convex bounding neighbourhood

引用

5th International Conference on Machine Learning and Cybernetics

作者： Nguyen, D. H. M. Wong, K. P. Chung, C. Y. Murdoch Univ Sch Engn Sci Murdoch WA 6150 Australia Hong Kong Polytech Univ Dept Elect Engn Computat Intelligence Applicat Res Lab Hong Kong Hong Kong Peoples R China

ISBN: (纸本)1424400600

This paper presents a new dynamic method of subpopulation in solving multi-modal search problems with evolutionary algorithms. The new method identify the modes found at each generation and equalises the subpopulation sizes assigned to each mode. Modes are identified sequentially starting with the highest fitness mode. Mode membership is determined by successive grouping of fitness dominated convex bounding neighbours, starting from the fittest individual. This new dynamic modal subpopulation approach is able to find a representative sample of optima for multi-modal landscape with infinite number of global and local optima with uneven heights and non-uniform distribution. The algorithm also facilitates parallel implementation.

关键词： multi-modal search evolutionary computation parallel algorithm subpopulation techniques

来源：评论

学校读者我要写书评

暂无评论

A Highly parallelizable Bond Fluctuation Model on the Body-Centered Cubic Lattice 11th

A Highly Parallelizable Bond Fluctuation Model on the Body-C...

引用

11th International Conference on parallel Processing and Applied Mathematics (PPAM)

作者： Jentzsch, Christoph Dockhorn, Ron Sommer, Jens-Uwe Leibniz Inst Polymer Res Dresden Dresden Germany Tech Univ Dresden Inst Theoret Phys Dresden Germany

ISBN: (纸本)9783319321523;9783319321516

We present a new Monte Carlo method which is based on the original Bond Fluctuation Model (scBFM) for simulating polymeric systems in three dimensions. A body centered cubic lattice is used instead of a simple cubic lattice. This modified Bond Fluctuation Model (bccBFM) fulfills the same requirements as the original scBFM, namely excluded volume and the cut-avoidance of bond vectors. Most remarkably the algorithm allows for a very efficient parallelization. This leads to a performance gain of about two orders of magnitude, when using graphics processor units (GPU). The bccBFM shows universal behavior both for static and dynamic properties and can be used to solve the same problems as the original scBFM, but provides an efficient implementation especially on GPUs.

关键词： Polymers Monte Carlo CUDA Bond fluctuation model parallel algorithm GPU

来源：评论

学校读者我要写书评

暂无评论

FPGA-accelerated Design of Motion Estimation for H.264 HDTV

FPGA-accelerated Design of Motion Estimation for H.264 HDTV

引用

2nd International Conference on Information and Computing Science

作者： Wang, Jintao Lu, Rui Qiu, Chu Gao, Pengdong Lu, Yongquan Yu, Wenhua Commun Univ China HPCC Beijing 100024 Peoples R China Electromagnet Commun Lab University Pk PA 16802 USA

ISBN: (纸本)9780769536347

H.264/AVC is a merging popular video coding standard which have new features. A H.264 open source, X264 code set, adopt motion estimation which consists of diamond search, hexagonal search, uneven multi-hexagon search, exhaustive search and hadamard exhaustive search. According to statistic values, the time-consuming of ME operations during HDTV(1920x1080p) encoding is about 60-80 percent within whole H.264 algorithm[1]. Considering the characteristics of FPGAs, flexibility, inherent parallel characteristic and the rapid prototyping, a parallel operation architecture of FPGA-based was designed with a Lattice ECP2M chip. The experiment results which include the time-consuming, the wave forms and the calculation cycles were analyzed with special tools, ispLever 6.1, ModelSim, etc. and the Post-Place & Route layout of a project was shown in the end of this paper. It is concluded that the FPGA chip can deal with HDTV sequences as a co-processor in a video coding system.

关键词： motion estimation FPGA-based co-processor parallel algorithm Pipeline Operation Block

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：