Allocating data to multicomputer nodes by physical optimization algorithms for loosely synchronous computations

Three optimization methods derived from natural sciences are considered for allocating data to multicomputer nodes. These are simulated annealing, genetic algorithms and neural networks. A number of design choices and the addition of preprocessing and postprocessing steps lead to versions of the algorithms which differ in solution qualities and execution times. In this paper the performances of these versions are critically evaluated and compared for test cases with different features. The performance criteria are solution quality, execution time, robustness, bias and parallelizability. Experimental results show that the physical algorithms produce better solutions than those of recursive bisection methods and that they have diverse properties. Hence, different algorithms would be suitable for different applications. For example, the annealing and genetic algorithms produce better solutions and do not show a bias towards particular problem structures, but they are slower than the neural network algorithms. Preprocessing graph contraction is one of the additional steps suggested for the physical methods. It produces a significant reduction in execution time, which is necessary for their applicability to large problems.     

用改进的遗传算法训练神经网络构造分类器

针对基本遗传算法存在容易早熟和局部搜索能力弱等缺陷,提出了改进的遗传算法,引入交叉概率和变异概率与个体的适度值相联系,改进了操作算子,而且在交叉操作后又引入模拟退火机制,提高遗传算法的局部搜索能力。同时,用改进的遗传算法和基本的遗传算法训练神经网络构造分类器,实验结果表明,改进的遗传算法在最好个体适度值和最好分类准确性等方面性能更好。     

利用遗传模拟退火算法优化神经网络结构

常用的神经网络是通过固定的网络结构得到最优权值,使网络的实用性受到影响。引入了一种基于方向的交叉算子和变异算子,同时把模拟退火算法引入了遗传算法,结合遗传算法和模拟退火算法的优点,提出了一种优化神经网络结构的遗传——模拟退火混合算法,实现了网络结构和权值的同时优化。仿真实验表明,与遗传算法和模拟退火算法相比,该算法优化的神经网络收敛速度较快、预测精度较高,提高了网络的处理能力。     

Application of genetic algorithms to optimization of neuronet recognition devices

The possibilities of applying genetic algorithms to optimization of the structure of neural networks that solve problems of recognition of handwritten and printed symbols and words are considered. The results of an experimental study are given. The experiments performed demonstrate an increase in the efficiency of neural networks after optimization. Ways of improving the results obtained are discussed. These results were partially obtained due to grant U4M000 of the Soros International Scientific Fund and also due to the“Neurocomputer” project (1992-994) of the State Committee of NAS of Ukraine on Science and Engineering. Translated from Kibernetika i Sistemnyi Analiz, No. 5, pp. 23–32, September–October, 1999.     

数据立方体实体化的多目标遗传算法

多维数据实视图选择问题是一个NP完全问题。提出一种基于约束的多目标优化遗传算法,将查询代价和维护代价分开考虑,更有效地解决复杂的实视图选择问题。实验结果表明,该算法具有更好的性能,特别是在获得的Pareto前沿的分布性上。     

Hybrid meta-heuristic optimization algorithms for time-domain-constrained data clustering

This paper addresses the question of time-domain-constrained data clustering, a problem which deals with data labelled with the time they are obtained and imposing the condition that clusters need to be contiguous in time (the time-domain constraint). The objective is to obtain a partitioning of a multivariate time series into internally homogeneous segments with respect to a statistical model given in each cluster.In this paper, time-domain-constrained data clustering is formulated as an unrestricted bi-level optimization problem. The clustering problem is stated at the upper level model and at the lower level the statistical models are adjusted to the set of clusters determined in the upper level. This formulation is sufficiently general to allow these statistical models to be used as black boxes. A hybrid technique based on combining a generic population-based optimization algorithm and Nelder–Mead simplex search is used to solve the bi-level model.The capability of the proposed approach is illustrated using simulations of synthetic signals and a novel application for survival analysis. This application shows that the proposed methodology is a useful tool to detect changes in the hidden structure of historical data.Finally, the performance of the hybridizations of particle swarm optimization, genetic algorithms and simulated annealing with Nelder–Mead simplex search are tested on a pattern recognition problem of text identification.     

人工神经网络与改进遗传算法的协作求解

简要介绍了改进遗传算法求解问题的步骤以及解决实际问题的特点。为了利用改进遗传算法的优点,提高其收敛速度,提出改进遗传算法与人工神经网络（BP网络）利用神经网络的联想记忆、特征提取功能辅助遗传算法求解结构优化设计问题,以避免在遗传算法中所作的那些不必要的分析计算,从而节省了计算时间。最后通过算例证实,比简单遗传算法与人工神经网络协作计算时间减少约25%。     

模糊作业时间的拆卸线平衡Pareto多目标优化

针对实际拆卸作业的复杂性,建立了考虑模糊作业时间的多目标拆卸线平衡问题的数学模型,提出了一种基于Pareto解集的多目标遗传模拟退火算法进行求解。改进了模拟退火操作的Metropolis准则,使其能够求解多目标优化问题。采用拥挤距离评价非劣解的优劣,保留了优秀个体,并通过精英选择策略,将非劣解作为遗传操作的个体,引导算法向最优方向收敛。基于25项拆卸任务算例,通过与现有的单目标人工蜂群算法进行对比,验证了所提算法的有效性和优越性。最后将该算法应用于某打印机拆卸线实例中,求得8种可选平衡方案,实现了求解结果的多样性。     

The Slack Method: A New Method for Static Allocation of Hard Real-Time Tasks

This article presents and evaluates the Slack Method, a new constructive heuristic for the allocation (mapping) of periodic hard real-time tasks to multiprocessor or distributed systems. The Slack Method is based on task deadlines, in contrast with other constructive heuristics, such as List Processing. The presented evaluation shows that the Slack Method is superior to list-processing-based approaches with regard to both finding more feasible solutions as well as finding solutions with better objective function values.In a comparative survey we evaluate the Slack Method against several alternative allocation techniques. This includes comparisons with optimal algorithms, non-guided search heuristics (e.g. Simulated Annealing), and other constructive heuristics. The main practical result of the comparison is that a combination of non-guided search and constructive approaches is shown to perform better than either of them alone, especially when using the Slack Method.     

Generating test data for both path coverage and fault detection using genetic algorithms

The aim of software testing is to find faults in a program under test, so generating test data that can expose the faults of a program is very important. To date, current stud- ies on generating test data for path coverage do not perform well in detecting low probability faults on the covered path. The automatic generation of test data for both path coverage and fault detection using genetic algorithms is the focus of this study. To this end, the problem is first formulated as a bi-objective optimization problem with one constraint whose objectives are the number of faults detected in the traversed path and the risk level of these faults, and whose constraint is that the traversed path must be the target path. An evolution- ary algorithm is employed to solve the formulated model, and several types of fault detection methods are given. Finally, the proposed method is applied to several real-world programs, and compared with a random method and evolutionary opti- mization method in the following three aspects： the number of generations and the time consumption needed to generate desired test data, and the success rate of detecting faults. The experimental results confirm that the proposed method can effectively generate test data that not only traverse the target path but also detect faults lying in it.     

Generating test data for both path coverage and fault detection using genetic algorithms

The aim of software testing is to find faults in a program under test, so generating test data that can expose the faults of a program is very important. To date, current studies on generating test data for path coverage do not perform well in detecting low probability faults on the covered path. The automatic generation of test data for both path coverage and fault detection using genetic algorithms is the focus of this study. To this end, the problem is first formulated as a bi-objective optimization problem with one constraint whose objectives are the number of faults detected in the traversed path and the risk level of these faults, and whose constraint is that the traversed path must be the target path. An evolutionary algorithmis employed to solve the formulatedmodel, and several types of fault detectionmethods are given. Finally, the proposed method is applied to several real-world programs, and compared with a random method and evolutionary optimization method in the following three aspects: the number of generations and the time consumption needed to generate desired test data, and the success rate of detecting faults. The experimental results confirm that the proposed method can effectively generate test data that not only traverse the target path but also detect faults lying in it.     

A Review of the Application of Meta-Heuristic Algorithms to 2D Strip Packing Problems

This paper is a review of the approachesdeveloped to solve 2D packing problems withmeta-heuristic algorithms. As packing tasks arecombinatorial problems with very large searchspaces, the recent literature encourages theuse of meta-heuristic search methods, inparticular genetic algorithms. The objective ofthis paper is to present and categorise thesolution approaches in the literature for 2Dregular and irregular strip packing problems.The focus is hereby on the analysis of themethods involving genetic algorithms. Anoverview of the methods applying othermeta-heuristic algorithms including simulatedannealing, tabu search, and artificial neuralnetworks is also given.