Abstract not found

This paper presents an approach for mapping tasks optimal to hardware and software components in order to design a real-time system. The tasks are derived from an algorithm and are represented by a task-graph. The performance of the algorithm on the resulting real-time system will meet the specified timing constraints. Some of the hardware components are programmable and others are application specific hardware processors. We propose a powerful MILP (Mixed Integer Linear Program) model with and without functional pipelining. The efficiency of the method is demonstrated with practical examples. 1 Introduction One important task of a hardware/software codesign is to map different tasks of an algorithm onto hardware or software components. Some components are programmable and others are application specific hardware processors. The aim of the codesign is to design a heterogeneous and loosely coupled multiprocessor system, which violates no timing constraints while performing the underly...

Workflow scheduling is one of the key issues in the management of workflow execution. Scheduling is a process that maps and manages execution of inter-dependent tasks on distributed resources. It introduces allocating suitable resources to workflow tasks so that the execution can be completed to satisfy objective functions specified by users. Proper scheduling can have significant impact on the performance of the system. In this chapter, we investigate existing workflow scheduling algorithms developed and deployed by various Grid projects.

Stochastic search algorithms inspired by physical and biological systems are applied to the problem of learning directed graphical probability models in the presence of missing observations and hidden variables. For this class of problems, deterministic search algorithms tend to halt at local optima, requiring random restarts to obtain solutions of acceptable quality. We compare three stochastic search algorithms: a Metropolis-Hastings Sampler (MHS), an Evolutionary Algorithm (EA), and a new hybrid algorithm called Population Markov Chain Monte Carlo, or popMCMC. PopMCMC uses statistical information from a population of MHSs to inform the proposal distributions for individual samplers in the population. Experimental results show that popMCMC and EAs learn more efficiently than the MHS with no information exchange. Populations of MCMC samplers exhibit more diversity than populations evolving according to EAs not satisfying physics-inspired local reversibility conditions. KEY WORDS: Markov Chain Monte Carlo, Metropolis-Hastings Algorithm, Graphical Probabilistic Models, Bayesian Networks, Bayesian Learning, Evolutionary Algorithms Machine Learning MCMC Issue 1 5/16/01 1.

In this paper computational aspects of the mathematical modelling of dynamic system evolution have been considered as a problem in information theory. The construction of such models is treated as a decision making process with limited available information. The solution of the problem is associated with a computational model based on heuristics of a Markov Chain in a discrete space-time of events. A stable approximation of the chain has been derived and the limiting cases are discussed. An intrinsic interconnection of constructive, sequential, and evolutionary approaches in related optimization problems provides new challenges for future work. Key words: decision making with limited information, optimal control theory, hyperbolicity of dynamic rules, generalized dynamic systems, Markov Chain approximation. 1 Introduction Many mathematical problems in information theory and optimal control related to dynamic system studies can be formulated in the following generic form. A decision...

Simulated Annealing has proven to be... In this paper we give a new proof of the convergence of Simulated Annealing by applying results about rapidly mixing Markov chains. With this proof technique it is possible to obtain better bounds for the finite time behaviour of Simulated Annealing than previously known.

Clustering has many uses such as the generation of taxonomies and concept formation. It is essentially a search through a model space to maximise a given criterion. The criterion aims to guide the search to find models that are suitable for a purpose. The search's aim is to efficiently and consistently find the model that gives the optimal criterion value. Considerable research has occurred into the criteria to use but minimal research has studied how to best search the model space. We describe how we have used simulated annealing to search the model space to optimise the minimum message length criterion.

A model of visual search is presented where gaze shifts are driven by an hybrid deterministic/stochastic mechanism operating over a saliency field. Results of the simulations are compared with experimental data, and a notion of complexity is used to quantify the behaviour of the system in different conditions. Index Terms — Eye movements, random walk, active vision, information encoding

this paper an application of Simulated Annealing to the 3-coloring problem is considered. In contrast to many good empirical results we will show for a certain class of graphs that the expected first hitting time of a proper coloring, given an arbitrary cooling scheme, is of exponential size. These results are complementary to those in [13], where the convergence of Simulated Annealing to an optimal solution in exponential time is proved. 1 Introduction

For the past 50 years, we have based most (if not all) of the world’s computers on the von Neumann model, which Alan Turing’s theoretical model in turn inspired in the first half of the 20th century. 1,2 Although we see the von Neumann model’s influence on some of today’s high-performance computers, the principles the model espouses are not adequate for solving many problems of great theoretical and practical importance. 3 Generally, a von Neumann model must execute a precise algorithm that can manipulate accurate data. However, lots of problems cannot meet such conditions—for example, accurate data might not be available, or a fixed or static algorithm cannot capture the complexity of the problem under study. Computational models based on natural phenomena have gained popularity in recent years. The need to solve a wide range of formidable problems that the prevailing mode of thinking could not solve has prompted the move in this direction. 4 Several recent studies show that nature-inspired techniques can potentially solve a wide range of problems and influence future computer design. 5 Some of these techniques are now commonplace and accepted by the wider scientific community. They tend to excel where the knowledge space is ambiguous or