A fast training support vector machine using parallel sequential minimal optimization is presented in this paper. Up to now, sequential minimal optimization (SMO) is one of the major algorithms for training SVM, but i...
详细信息
ISBN:
(纸本)9780878492459
A fast training support vector machine using parallel sequential minimal optimization is presented in this paper. Up to now, sequential minimal optimization (SMO) is one of the major algorithms for training SVM, but it still requires a large amount of computation time for the large sample problems. Unlike the traditional SMO, the parallel SMO partitions the entire training data set into small subsets first and then runs multiple CPU processors to seal with each of the partitioned data set. Experiments show that the new algorithm has great advantage in terms of speediness when applied to problems with large training sets and high dimensional spaces without reducing generalization performance of SVM.
Time series motifs are an integral part of diverse data mining applications including classification, summarization and near-duplicate detection. These are used across wide variety of domains such as image processing,...
详细信息
ISBN:
(纸本)9783642152900
Time series motifs are an integral part of diverse data mining applications including classification, summarization and near-duplicate detection. These are used across wide variety of domains such as image processing, bioinformatics, medicine, extreme weather prediction, the analysis of web log and customer shopping sequences, the study of XML query access patterns, electroencephalograph interpretation and entomological telemetry data mining. Exact Motif discovery in soft real-time over 100K time series is a challenging problem. We present novel parallel algorithms for soft real-time exact motif discovery on multi-core architectures. Experimental results on large scale P6 SMP system, using real life and synthetic time series data, demonstrate the scalability of our algorithms and their ability to discover motifs in soft real-time. To the best of our knowledge, this is the first such work on parallel scalable soft real-time exact motif discovery.
Biosignal processing of large data sets such as electro- or magnetoencephalographic recordings can be computationally very demanding. The usual approach of using a high level language can be to slow especially in the ...
详细信息
ISBN:
(纸本)9783642038815
Biosignal processing of large data sets such as electro- or magnetoencephalographic recordings can be computationally very demanding. The usual approach of using a high level language can be to slow especially in the application of statistical methods such as independent component analysis. If heuristic parameter space searches or bootstrapping are required the computations exceed the capabilities of a standard desktop PC. With the advent of small PC clusters and multi-core CPUs parallel programming has become much more accessible. Here the parallel implementation of an independent component analysis algorithm is presented, which runs on simple multi processing hardware well below the level of high performance clusters. One feature of our approach is the close interaction between parallel program and high level language simplifying development and testing.
The paper introduced recursive algorithm of fractal graphics, put forward fractal graphics parallel algorithm. Analyzing recursive algorithmic time complexity and speedup rate of the parallel algorithm. The experiment...
详细信息
ISBN:
(纸本)9780769541105
The paper introduced recursive algorithm of fractal graphics, put forward fractal graphics parallel algorithm. Analyzing recursive algorithmic time complexity and speedup rate of the parallel algorithm. The experimental results of PC cluster show that the theoretical analysis and the experimental results of fractal graphics parallel algorithm are consistency with a marked speedup rate.
In this paper, we present a parallel mesh surface generation approach for unorganized point clouds that runs on the graphics processing unit (GPU) Our approach integrates point cloud simplification, point cloud optimi...
详细信息
ISBN:
(纸本)9780791848999
In this paper, we present a parallel mesh surface generation approach for unorganized point clouds that runs on the graphics processing unit (GPU) Our approach integrates point cloud simplification, point cloud optimization, and local triangulation techniques into the same framework The input point cloud will be processed through three steps of algorithms, which are 1) preprocessing to generate the neighborhood table of points and estimate the normal vectors, 2) clustering to group points into optimized clusters that minimize the shape approximation error, and 3) meshing to connect the seed points in clusters to form the resultant triangular mesh surface As the number of clusters can be specified by users, the number of vertices on resultant mesh surfaces is controlled The algorithms exploited here are highly parallelized to take advantage of the single-instruction-multiple-data (SIMD) parallelism that is available on consumer-level graphics hardware with GPU Moreover, to overcome memory limitation on graphics hardware the algorithms in all these steps are able to process massive data in streaming mode
A parallel evolutionary approach of Compaction Problem is introduced using Map Reduce. This problem is of interest for VLSI testing and bioinformatics. The overall cost of a VLSI circuit's testing depends on the l...
详细信息
ISBN:
(纸本)9783642158704
A parallel evolutionary approach of Compaction Problem is introduced using Map Reduce. This problem is of interest for VLSI testing and bioinformatics. The overall cost of a VLSI circuit's testing depends on the length of its test sequence;therefore the reduction of this sequence, keeping the coverage, will lead to a reduction of used resources in the testing process. The problem of finding minimal test sets is NP-hard. We introduce a distributed evolutionary algorithm (Map Reduce parallel Evolutionary algorithm-MRPEA) and compare it with two greedy approaches. The proposed algorithms are evaluated on randomly generated five-valued benchmarks that are scalable in size. The Map Reduce paradigm offers the possibility to distribute and scale large amount of data. Experiments show the efficiency of the proposed parallel approach. The project, containing the Hadoop implementation can be found at: http://***/projects/dcpsolver/[10].
We describe a class of adaptive algorithms tor approximating the global minimum of a function defined on a compact subset of R(d) The algorithms are adaptive versions of Monte Carlo search and use a memory of a fixed ...
详细信息
We describe a class of adaptive algorithms tor approximating the global minimum of a function defined on a compact subset of R(d) The algorithms are adaptive versions of Monte Carlo search and use a memory of a fixed number of past observations By choosing a large enough memory. the convergence rate can be made to exceed any power of the convergence rate obtained with standard Monte Carlo search (C) 2008 IMACS Published by Elsevier B V All rights reserved
Using highly effective parallel calculations, it is shown that in a moment elastic medium, there is a resonant frequency which corresponds to the eigenfrequency of rotational motion of particles and does not depend on...
详细信息
Using highly effective parallel calculations, it is shown that in a moment elastic medium, there is a resonant frequency which corresponds to the eigenfrequency of rotational motion of particles and does not depend on the size of the region studied.
We present a novel distributed algorithm for the maximal independent set problem (This is an extended journal version of Schneider and Wattenhofer in Twenty-seventh annual ACM SIGACT-SIGOPS symposium on principles of ...
详细信息
We present a novel distributed algorithm for the maximal independent set problem (This is an extended journal version of Schneider and Wattenhofer in Twenty-seventh annual ACM SIGACT-SIGOPS symposium on principles of distributed computing, 2008). On bounded-independence graphs our deterministic algorithm finishes in O(log* n) time, n being the number of nodes. In light of Linial's Omega(log* n) lower bound our algorithm is asymptotically optimal. Furthermore, it solves the connected dominating set problem for unit disk graphs in O(log* n) time, exponentially faster than the state-of-the-art algorithm. With a new extension our algorithm also computes a delta + 1 coloring and a maximal matching in O(log* n) time, where delta is the maximum degree of the graph.
＜正＞While high throughput screening（HTS） and combinational chemistry（CC） become mature, the need of processing jillion chemical structure data is growing *** is a great challenge to a chemical database search en...
详细信息
＜正＞While high throughput screening（HTS） and combinational chemistry（CC） become mature, the need of processing jillion chemical structure data is growing *** is a great challenge to a chemical database search engine when it searches on a database with millions of chemical structures. Most of chemical substructure search engines were published before the year *** recent years, high performance computing and personal computing have experienced dramatic improvements.
暂无评论