In this paper we propose a new parallel algorithm for mining association rules. In comparison to the previously proposed parallelalgorithmsthat are based on Apriori, our algorithm is effective with much smaller memo...
详细信息
ISBN:
(纸本)9781581130614
In this paper we propose a new parallel algorithm for mining association rules. In comparison to the previously proposed parallelalgorithmsthat are based on Apriori, our algorithm is effective with much smaller memory space, and does not require any redundancy for transferring, loading neither processing replicated data. Our algorithm handles the balancing of the load among the processors by dynamically employing a skew resolution and two skew avoidance methods. Performance evaluation of the implementation of our algorithm in a shared-nothing system shows that our skew handling techniques avoid the disastrous performance degradation that a parallel algorithm with no load balance suffers on skewed data, and shows the same performance in the case of non-skewed data. Copyright ACM 1998.
this paper discusses the main achievements of the EPIC project, whose aim was to design a high level programming environment with an associated implementation for portable parallel image processing. the project was fu...
详细信息
ISBN:
(纸本)3540649522
this paper discusses the main achievements of the EPIC project, whose aim was to design a high level programming environment with an associated implementation for portable parallel image processing. the project was funded as part of the EPSRC Portable Software Tools for parallelarchitectures (PSTPA) programme. the paper summarises new portable programming abstractions for image processing, and outlines the automatically optimising implementation which achieves portability of application code and efficiency of implementation on a closely coupled distributed memory parallel system. the paper includes timings for optimised and unoptimised versions of typical image processingalgorithms;it draws the main conclusion that it is possible to achieve portability with efficiency, for a specific application, by adopting a high level algebraic programming model, together with a transformation-based optimiser which reclaims the loss of efficiency which an algebraic approach traditionally entails.
A concept for a future integer arithmetic unit as well as a first implementation of the arithmetic unit's core as smart pixel detector chip is presented. this architecture is well-suited for a realization with 3-D...
详细信息
ISBN:
(纸本)0818685727
A concept for a future integer arithmetic unit as well as a first implementation of the arithmetic unit's core as smart pixel detector chip is presented. this architecture is well-suited for a realization with 3-D optoelectronic very large scale integrated (VLSI) circuits. Due to the use of optical interconnections running vertically to the circuit's surface no pin limitation is given. this allows massively parallelism and a higher throughput performance than in all-electronic solutions. To exploit the potential of optical interconnections in VLSI systems efficiently well-adapted low-level algorithms and architectures have to be developed. this is demonstrated for a pipelined arithmetic unit using a redundant number representation. A gate layout for the optoelectronic circuits is given as well as a specification for the necessary optical interconnection scheme linking the circuits with free-space optics. It is shown that the throughput can be increased by a factor of 10 to 50 compared to current all-electronic processors by considering state-of-the-art optical and optoelectronic technolgy.
High Performance Fortran (HPF) is the de facto standard language for writing data parallel programs. In case of applications that use indirect addressing on distributed arrays, HPF compilers have limited capabilities ...
详细信息
ISBN:
(纸本)3540649522
High Performance Fortran (HPF) is the de facto standard language for writing data parallel programs. In case of applications that use indirect addressing on distributed arrays, HPF compilers have limited capabilities for optimizing such codes on distributed memory architectures, especially for optimizing communication and reusing communication schedules between subroutine boundaries. this paper describes a dynamic approach for optimizing unstructured communication in codes with indirect addressing. the basic idea is that runtime data reflecting the communication patterns will be reused if possible. the user has only to specify which data in the program has to be traced for modifications. the experiments and results show the effectiveness of the chosen approach.
this paper presents algorithms and architectures for implementing from 1-D to multidimensional M-D digital nonrecursive filters. these architectures are very regular and support single chip implementation in VLSI, as ...
详细信息
the proceedings contain 79 papers. the special focus in this conference is on Economics, Emergence, Complex Systems, Issues and Innovations in Evolutionary Computation. the topics include: A framework for evolutionary...
ISBN:
(纸本)3540648917
the proceedings contain 79 papers. the special focus in this conference is on Economics, Emergence, Complex Systems, Issues and Innovations in Evolutionary Computation. the topics include: A framework for evolutionary processes in semantic memory;preferential partner selection in evolutionary labor markets;subspace pursuit for exploratory modeling;evolutionary computing in multi-agent environments;evalution of a simple host-parasite genetic algorithm;testing three paradigms for evolving groups of cooperative, simple agents;acquisition of general adaptive features by evolution;hybrid interior-lagrangian penalty based evolutionary optimization;GA-optimal fitness functions;scaling up evolutionary programming algorithms;short notes on the schema theorem and the building block hypothesis in genetic algorithms;a superior evolutionary algorithm for 3-SAT;evolvable hardware control for dynamic reconfigurable and adaptive computing;evolutionary programming strategies with self-adaptation applied to the design of rotorcraft using parallelprocessing;optimization of thinned phased arrays using evolutionary programming;evolutionary domain covering of an inference system for function approximation;learning to re-engineer semantic networks using cultural algorithms;integration of slicing methods into a cultural algorithm in order to assist in large-scale engineering systems design;genetic search for object identification;fuzzy cultural algorithms with evolutionary programming;culturing evolution strategies to support the exploration of novel environments by an intelligent robotic agent;skeuomorphs and cultural algorithms;sphere operators and their applicability for constrained parameter optimization problems;numeric mutation as an improvement to symbolic regression in genetic programming and variable-dimensional optimization with evolutionary algorithms using fixed-length representations.
the structural specification and modeling of time critical real-time systems has become a major area for recent research topics. this is particularly relevant for computer music when sound computation is realized invo...
详细信息
Data-accumulating algorithms for sorting was investigated. For the study a bound for the running time was obtained and from the result, sorting d-algorithms was characterized in terms of complexity. It was shown that,...
详细信息
Data-accumulating algorithms for sorting was investigated. For the study a bound for the running time was obtained and from the result, sorting d-algorithms was characterized in terms of complexity. It was shown that, if the data arrive too fast, then a successful algorithm must have a running time upper bounded by a constant. the results also show the importance of parallelism for the theory of d-algorithms.
the use of PC clusters interconnected by high performance local networks is one of the major current trends in parallel/distributed computing. We give coarse-grained, BSP-like, parallelalgorithms to solve many proble...
详细信息
ISBN:
(纸本)3540649522
the use of PC clusters interconnected by high performance local networks is one of the major current trends in parallel/distributed computing. We give coarse-grained, BSP-like, parallelalgorithms to solve many problems arising in the context of interval graphs, namely connected components, maximum weighted clique, BFS and DFS trees, minimum interval covering, maximum independent set and minimum dominating set. All of the described p-processor parallelalgorithms require only constant or O(log p) number of communication rounds and are efficient in practice, as demonstrated by our experimental results obtained on a Fast Ethernet based PC cluster.
A lattice-gas Automata two-dimensional program was developed for analysis of single and two-phase flow behaviors, to support the development of integrated software modules for Nuclear Power Plant mechanistic simulatio...
详细信息
A lattice-gas Automata two-dimensional program was developed for analysis of single and two-phase flow behaviors, to support the development of integrated software modules for Nuclear Power Plant mechanistic simulations. the program has single-color, which includes FHP I, II, and III models, two-color (Immiscible lattice gas), and two-velocity methods including a gravity effect model. Parameter surveys have been performed for Karman vortex street, two-phase separation for understanding flow regimes, and natural circulation flow for demonstrating passive reactor safety due to the chimney structure vessel. In addition, lattice-Boltzmann Equation two-dimensional programs were also developed. For analyzing single-phase flow behavior, a lattice-Boltzmann-BGK program was developed, which has multi-block treatments. A Finite Differential lattice-Boltzmann Equation program of parallelized version was introduced to analyze boiling two-phase flow behaviors. Parameter surveys have been performed for backward facing Bow, Karman vortex street, bent piping flow with/without obstacles for piping system applications, flow in the porous media for demonstrating porous debris coolability, Couette Row, and spinodal decomposition to understand basic phase separation mechanisms. parallelization was completed by using a domain decomposition method for all of the programs. An increase in calculation speed of at least 25 times, by parallelprocessing on 32 processors, demonstrated high parallelization efficiency. Application fields for microscopic model simulation to hypothetical severe conditions in large plants were also discussed.
暂无评论