In this paper we discuss the effect of two tree application parameters, namely computational granularity and tree imbalance, on the overall performance of parallel tree structured application. We apply a technique cal...
详细信息
In this paper we discuss the effect of two tree application parameters, namely computational granularity and tree imbalance, on the overall performance of parallel tree structured application. We apply a technique called phased-based adaptive dynamic load balancing, a version of adaptive DLB in which a parallel computation moves through different load balancing phases identified on the basis of run-time workloads. A simulator was used to evaluate the benefit of this approach. Two DLB algorithms were used as test cases;the Generalized Dimension Exchange Method and Load server algorithms. The results show that applications with large grain computation would benefit most from our approach while tree imbalance does not show consistent advantage.
The present study addresses the influence of acoustic source excitations to the dynamic of structures. A particular example being dealt with is the coupled acoustically modified aeroelastic instability problem. To thi...
详细信息
The present study addresses the influence of acoustic source excitations to the dynamic of structures. A particular example being dealt with is the coupled acoustically modified aeroelastic instability problem. To this end, as a first step, the boundary element formulation of the governing Helmholtz equation for the acoustic wave propagation in fluids is set up, and the computational scheme is worked out to obtain the pressure field. Further, reflection and absorption in the more complex scattering phenomena have been disregarded in the treatment of the acoustic wave propagation problem. Hence the acoustic terms on the right hand side can be calculated as an isolated problem using Boundary Element Method. As a first approximation, the applicability of the principles of superposition is assumed, and the pressure field due to the acoustic wave propagation is incorporated as excitation force field in the Structural Dynamic equation of structures. The solution to this equation is worked out for a coupled unsteady aerodynamic and acoustic excitation in the aeroelastic instability problem. A simple beam example for such acousto-aeroelastic problem is then worked out and has given promising result. The results are validated by resort to earlier work for similar cases.
It is well-known that Hebbian synapses, with appropriate weight normalization, extract the first principal component of the input patterns. Anti-Hebb rules have been used in combination with Hebb rules to extract addi...
详细信息
ISBN:
(纸本)0780341228
It is well-known that Hebbian synapses, with appropriate weight normalization, extract the first principal component of the input patterns. Anti-Hebb rules have been used in combination with Hebb rules to extract additional principal components or generate sparse codes. Here we show that the simple anti-Hebbian synapses alone can support an important computational function: solving simultaneous linear equations. During repetitive learning with a simple anti-Hebb rule, the weights onto an output unit always converge to the exact solution of the linear equations whose coefficients correspond to the input patterns and whose constant terms correspond to the biases, provided that the solution exists. If there are more equations than unknowns and no solution exists, the weights approach the values obtained by using the Moore-Penrose generalized inverse (pseudoinverse). No explicit matrix inversion is involved and there is no need to normalize weights. Mathematically, the anti-Hebb rule may be regarded as an iterative algorithm for learning a special case of the linear associative mapping. Since solving systems of linear equations is a very basic computational problem to which many other problems are often reduced, our interpretation suggests a potentially general computational role for the anti-Hebbian synapses and a certain type of long-term depression (LTD).
Computation in biology and in conventional computer architectures seem to share some features, yet many of their important characteristics are very different. To address this, [1] introduced systemic computation, a mo...
详细信息
ISBN:
(纸本)159593698X
Computation in biology and in conventional computer architectures seem to share some features, yet many of their important characteristics are very different. To address this, [1] introduced systemic computation, a model of interacting systems with natural characteristics. Following this work, here we introduce the first platform implementing such computation, including programming language, compiler and virtual machine. To investigate their use we then provide an implementation of a genetic algorithm applied to the travelling salesman problem and also explore how SC enables self-adaptation with the minimum of additional code. Copyright 2007 ACM.
This paper proposes a method of describing narrative structure, that focuses on the behavior of the characters in the story. It also proposes to assign the concepts of focus, polarity, dynamic, motivation, and result ...
详细信息
Computations of a lexicographic ordering for subgroups and cosets of permutation groups were presented. Computations of finite groups with a faithful permutation representation were also discussed. The algorithms were...
详细信息
ISBN:
(纸本)9781581136418
Computations of a lexicographic ordering for subgroups and cosets of permutation groups were presented. Computations of finite groups with a faithful permutation representation were also discussed. The algorithms were implemented in the GAP (Groups, Algorithms, and Programming) group. The extension to more general settings is the novel feature of the computations.
Though humpback whales have been observed accurately locating singing whales from several kilometers distance, little is known about the sound localization cues whales use. Humpback whales might use binaural cues such...
详细信息
ISBN:
(纸本)9781622764655
Though humpback whales have been observed accurately locating singing whales from several kilometers distance, little is known about the sound localization cues whales use. Humpback whales might use binaural cues such as differences in time of sound arrival and differences in received intensity to locate singers in azimuth. However, it is not clear that this is feasible because the differences between songs received at each of a humpback whale's ears should be quite small due to the speed of sound in water and the relatively close positioning of the whales inner ears. In this study, sound localization in whales is explored though the use of computational models. Synthesized sounds offset to represent interaural timing cues (rITDs) or filtered to represent interaural level cues (rILDs) were created from pure tones and compared using neural networks, computational models that are designed to model neuronal function. The neural networks' performances were found to accurately reflect the localization limitations of biological systems. Results revealed several potential cues for localization in azimuth. Interaural timing differences were most useful between 0.5-1.75 kHz and accuracy was greater at the midline, decreasing toward the periphery. Spectral cues also changed systematically with angle with lower boundaries between 0.15-0.75 kHz for sounds representing ILDs produced by spherical objects with a diameter of 1.0 and 0.3 m. These diameters were chosen to represent the approximate diameter of a whale's skull and the distance between a whale's inner ears respectfully. Despite the challenges of localizing sound in v/ater, cues similar to those used by terrestrial mammals appear to be available to whales.
When fitting simultaneously-active dipoles, currently-used DSL (Dipole Source Localization) methods may contain MulGenErrs (Multiple-Generator Errors) due to model misspecification. The size of the MulGenErr is a func...
详细信息
When fitting simultaneously-active dipoles, currently-used DSL (Dipole Source Localization) methods may contain MulGenErrs (Multiple-Generator Errors) due to model misspecification. The size of the MulGenErr is a function of both the model used, and the dipole parameters, including the dipoles' waveforms (time-varying magnitudes). For a given fitting model, by examining the variation of the MulGenErrs (or the fit parameters) under different waveforms for the same generating-dipoles, the accuracy of the fitting model for this set of dipoles can be determined. This method of testing model misspecification can be applied to evoked potential maps where the parameters of the generating-dipoles are unknown. The dipole parameters fitted in a model should only be accepted if the model can be shown to be sufficiently accurate.
The availability of computationally efficient and accurate methods for probabilistic computation is crucial to the success of applications of probabilistic design using complex engineering simulation models. To addres...
详细信息
The availability of computationally efficient and accurate methods for probabilistic computation is crucial to the success of applications of probabilistic design using complex engineering simulation models. To address this need, a Saddlepoint Approximation method for probabilistic engineering analysis is introduced. A general performance function is approximated at the Most Likelihood Point with either linear or quadratic forms and the Saddlepoint Approximation is then applied to evaluate the probability associated with the performance. The proposed approach provides highly accurate probabilistic results while maintaining minimum computational requirement. Two examples are presented to demonstrate the effectiveness of the proposed method.
The paper explores students' learning about gases in chemistry through constructing computational models of complex systems with the new *** platform (MMM;[7]). The design of MMM is based on the agent-based modell...
详细信息
暂无评论