A heterogeneous cooperative parallel search of branch-and-bound method and tabu search algorithm

In this study, we present a heterogeneous cooperative parallel search that integrates branch-and-bound method and tabu search algorithm. These two algorithms perform searches in parallel and cooperate by asynchronously exchanging information about the best solutions found and new initial solutions for tabu search. The rapid production of a good solution from the tabu search process provides the branch-and-bound process with a better feasible solution to accelerate the elimination of subproblems that do not contain an optimal solution. The new initial solution produced from the subproblem with a least-cost lower bound of the branch-and-bound method suggests the best potential area for tabu search to explore. We use a master-slave model to reduce the complexity of communication and enhance the performance of data exchange. A branch-and-bound process is used as the master process to control the exchange of information and the termination of computation. Several tabu search processes are executed simultaneously as the slave processes and cooperate by asynchronously exchanging information on the best solutions found and the new initial solutions by the master process of branch-and-bound. Based on the computation experiments of solving traveling salesman problems (TSP), the proposed heterogeneous parallel search algorithm outperforms a conventional parallel branch-and-bound method and a conventional parallel tabu search. We also present the computational results showing the efficiency of heterogeneous cooperative parallel search when we use more processors to accelerate search time. Thus, the proposed heterogeneous parallel search algorithm achieves linear accelerations.     

A genetic algorithm with tournament selection as a local search method

Some sufficient conditions are found under which the generational genetic algorithm with tournament selection first visits a local optimum in polynomially bounded time on average. These conditions are satisfied on a class of problems with guaranteed local optima if the appropriate parameters of the algorithm are chosen.     

An adaptive search algorithm for numerical optimization

The flexible polyhedron (simplex) search algorithm is reviewed and some of its shortcomings highlighted. Particularly, the fixed search parameters are shown to be a sure liability and an improvement is proposed. A unidirectional optimal search algorithm is substituted for the set of fixed rules usually employed to modify the simplex. This modification proves especially effective in dealing with “narrow valley” situations, normally encountered whenever the decision variables exhibit some degree of correlation. The new adaptive algorithm compares well with the parent simplex method, featuring less function evaluations and better convergence properties in cases where the classical search techniques perform poorly or fail altogether.     

A parallel search algorithm for directed acyclic graphs

A parallel algorithm for depth-first searching of a directed acyclic graph (DAG) on a shared memory model of a SIMD computer is proposed. The algorithm uses two parallel tree traversal algorithms, one for the preorder traversal and the other for therpostorder traversal of an ordered tree. Each of these traversal algorithms has a time complexity ofO(logn) whenO(n) processors are used,n being the number of vertices in the tree. The parallel depth-first search algorithm for a directed acyclic graphG withn vertices has a time complexity ofO((logn)²) whenO(n ^2.81/logn) processors are used.     

A stochastic local search algorithm with adaptive acceptance for high-school timetabling

Automating high school timetabling is a challenging task. This problem is a well known hard computational problem which has been of interest to practitioners as well as researchers. High schools need to timetable their regular activities once per year, or even more frequently. The exact solvers might fail to find a solution for a given instance of the problem. A selection hyper-heuristic can be defined as an easy-to-implement, easy-to-maintain and effective ‘heuristic to choose heuristics’ to solve such computationally hard problems. This paper describes the approach of the team hyper-heuristic search strategies and timetabling (HySST) to high school timetabling which competed in all three rounds of the third international timetabling competition. HySST generated the best new solutions for three given instances in Round 1 and gained the second place in Rounds 2 and 3. It achieved this by using a fairly standard stochastic search method but significantly enhanced by a selection hyper-heuristic with an adaptive acceptance mechanism.     

基于梯度的自适应快速布谷鸟搜索算法

针对标准布谷鸟搜索(CS)算法存在全局搜索和局部搜索能力不平衡的缺点, 提出一种基于梯度的自适应快速布谷鸟搜索(GBAQCS)算法. 在改进的算法中, 针对偏好随机游动的步长, 在利用目标函数的梯度决定步长方向的基础上, 首先提出自适应搜索机制平衡了算法的全局搜索和局部搜索能力; 其次提出快速 搜索策略, 充分利用当前鸟巢信息进行精细化搜索, 从而提高算法的搜索精度和收敛速度. 实验结果表明, 相比其他算法, 所提出的改进策略使算法的全局搜索和局部搜索能力保持了相对的平衡, 并提高了算法的收敛性能.     

The island confinement method for reducing search space in local search methods

Typically local search methods for solving constraint satisfaction problems such as GSAT, WalkSAT, DLM, and ESG treat the problem as an optimisation problem. Each constraint contributes part of a penalty function in assessing trial valuations. Local search examines the neighbours of the current valuation, using the penalty function to determine a “better” neighbour valuation to move to, until finally a solution which satisfies all the constraints is found. In this paper we investigate using some of the constraints as “hard” constraints, that are always satisfied by every trial valuation visited, rather than as part of a penalty function. In this way these constraints reduce the possible neighbours in each move and also the overall search space. The treating of some constraints as hard requires that the space of valuations that are satisfied is “connected” in order to guarantee that a solution can be found from any starting position within the region; thus the concept of islands and the name “island confinement method” arises. Treating some constraints as hard provides new difficulties for the search mechanism since the search space becomes more jagged, and there are more deep local minima. A new escape strategy is needed. To demonstrate the feasibility and generality of our approach, we show how the island confinement method can be incorporated in, and significantly improve, the search performance of two successful local search procedures, DLM and ESG, on SAT problems arising from binary CSPs. A preliminary version of this paper appeared in AAAI’2002.     

A parallel adaptive grid algorithm for computational shock hydrodynamics

Adaptive Mesh Refinement (AMR) algorithms that dynamically match the local resolution of the computational grid to the numerical solution being sought have emerged as powerful tools for solving problems that contain disparate physical scales. In particular several workers have demonstrated the effectiveness of employing an adaptive, hierarchical block-structured grid system for time-accurate simulations of complex shock wave phenomena. Here we present an overview of one such block-structured AMR algorithm. Our formulation has progressed far beyond the development stage to become a reliable numerical tool for performing detailed investigations of complex flows. While our refinement machinery is not tied to a specific application and is intended for general use, in this paper we adopt detonation phenomena as a theme so as to provide a sense of purpose.     

A parallel genetic algorithm for optimization of trading strategies

An approach to optimization of trading strategies (algorithms) is described based on indicators of financial markets and on evolutionary computations. A parallel genetic algorithm is presented for searching optimal parameters of trading strategies aimed at profit maximization.     

A hybrid genetic algorithm for multiobjective problems with activity analysis-based local search

The objective of this research was the development of a method that integrated an activity analysis model of profits from production with a biophysical model, and included the capacity for optimization over multiple objectives. We specified a hybrid genetic algorithm using activity analysis as a local search method, and NSGA-II for calculation of the multiple objective Pareto optimal set. We describe a parallel computing approach to computation of the genetic algorithm, and apply the algorithm to evaluation of an input tax to regulate pollution from agricultural production.     

A cooperative parallel tabu search algorithm for the quadratic assignment problem

In this study, we introduce a cooperative parallel tabu search algorithm (CPTS) for the quadratic assignment problem (QAP). The QAP is an NP-hard combinatorial optimization problem that is widely acknowledged to be computationally demanding. These characteristics make the QAP an ideal candidate for parallel solution techniques. CPTS is a cooperative parallel algorithm in which the processors exchange information throughout the run of the algorithm as opposed to independent concurrent search strategies that aggregate data only at the end of execution. CPTS accomplishes this cooperation by maintaining a global reference set which uses the information exchange to promote both intensification and strategic diversification in a parallel environment. This study demonstrates the benefits that may be obtained from parallel computing in terms of solution quality, computational time and algorithmic flexibility. A set of 41 test problems obtained from QAPLIB were used to analyze the quality of the CPTS algorithm. Additionally, we report results for 60 difficult new test instances. The CPTS algorithm is shown to provide good solution quality for all problems in acceptable computational times. Out of the 41 test instances obtained from QAPLIB, CPTS is shown to meet or exceed the average solution quality of many of the best sequential and parallel approaches from the literature on all but six problems, whereas no other leading method exhibits a performance that is superior to this.     

Non-revisiting genetic algorithm with adaptive mutation using constant memory

The continuous non-revisiting genetic algorithm (cNrGA) uses the entire search history and parameter-less adaptive mutation to significantly enhance search performance. Storing the entire search history is natural and costs little when the number of fitness evaluations is small or moderate. However, if the number of evaluations required is substantial, some memory management is desirable. In this paper, we propose two pruning mechanisms to keep the memory used constant. They are least recently used pruning and random pruning. The basic idea is to prune a unit of memory when the memory threshold is reached and some new search information is required to be stored, thus keeping the overall memory used constant. Meanwhile, both pruning strategies naturally form parameter-less adaptive mutation operators. A study is carried out to evaluate the impact on performance caused by loss of search history information. Experimental results show that (1) both strategies can maintain the performance of cNrGA, up to the empirical limit when 90 % of the search history is not recorded, (2) cNrGA and its variants with constant memory outperform the real-coded genetic algorithm and the standard particle swarm optimization. By pre-extracting all the current prune-able history information and storing them into a list, namely, to-prune-list, the overhead of both pruning strategies becomes small. This suggests that cNrGA can be extended to use in situations when the number of fitness evaluations is much larger than before with no significant effect on statistical performance. This widens the applicability of cNrGA to include more practical problems that require larger number of fitness evaluations before converging to the global optima.     

A global optimization algorithm using adaptive random search

A new random-search global optimization is described in which the variance of the step-size distribution is periodically optimized. By searching over a variance range of 8 to 10 decades, the algorithm finds the step-size distribution that yields the best local improvement in the criterion function. The variance search is then followed by a specified number of iterations of local random search where the step-size variance remains fixed. Periodic wide-range searches are introduced to ensure that the process does not stop at a local minimum. The sensitivity of the complete algorithm to various search parameters is investigated experimentally for a specific test problem. The ability of the method to locate global minima is illustrated by an example. The method also displays considerable problem independence, as demonstrated by two large and realistic example problems: (1) the identification of 25 parameters in a nonlinear model of a five-degrees-of-freedom mechanical dynamic system and (2) solution of a 24-parameter inverse problem required to identify a pulse train whose frequency spectrum matched a desired reference spectrum.     

Selecting informative rules with parallel genetic algorithm in classification problem

The classification system is very important for making decision and it has been attracted much attention of many researchers. Usually, the traditional classifiers are either domain specific or produce unsatisfactory results over classification problems with larger size and imbalanced data. Hence, genetic algorithms (GA) are recently being combined with traditional classifiers to find useful knowledge for making decision. Although, the main concerns of such GA-based system are the coverage of less search space and increase of computational cost with the growth of population. In this paper, a rule-based knowledge discovery model, combining C4.5 (a Decision Tree based rule inductive algorithm) and a new parallel genetic algorithm based on the idea of massive parallelism, is introduced. The prime goal of the model is to produce a compact set of informative rules from any kind of classification problem. More specifically, the proposed model receives a base method C4.5 to generate rules which are then refined by our proposed parallel GA. The strength of the developed system has been compared with pure C4.5 as well as the hybrid system (C4.5 + sequential genetic algorithm) on six real world benchmark data sets collected from UCI (University of California at Irvine) machine learning repository. Experiments on data sets validate the effectiveness of the new model. The presented results especially indicate that the model is powerful for volumetric data set.     

A dynamic adaptive local search algorithm for the circular packing problem

This paper studies the circular packing problem (CPP) which consists of packing n non-identical circles C_i of known radius r_i, i ∈ N = {1, … , n}, into the smallest containing circle C. The objective is to determine the coordinates (x_i, y_i) of the center of C_i, i ∈ N, as well as the radius r and center (x, y) of C. This problem, which is a variant of the two-dimensional open dimension problem, is solved using a two-step, dynamic, adaptive, local search algorithm. At each iteration, the algorithm identifies the set of potential “best local positions” of a circle C_i, i ∈ N, given the positions of the previously packed circles, and determines for each of these positions the coordinates and radius of the smallest containing circle. The “best local position” minimizes the radius of the current containing circle. That is, every time an additional circle is packed, both the center and the radius of the containing circle are dynamically updated, and the smallest containing circle is known. The experimental results reflect the good performance of the algorithm.     

Parameter adaptive harmony search algorithm for unimodal and multimodal optimization problems

This paper presents a parameter adaptive harmony search algorithm (PAHS) for solving optimization problems. The two important parameters of harmony search algorithm namely Harmony Memory Consideration Rate (HMCR) and Pitch Adjusting Rate (PAR), which were either kept constant or the PAR value was dynamically changed while still keeping HMCR fixed, as observed from literature, are both being allowed to change dynamically in this proposed PAHS. This change in the parameters has been done to get the global optimal solution. Four different cases of linear and exponential changes have been explored. The change has been allowed during the process of improvization. The proposed algorithm is evaluated on 15 standard benchmark functions of various characteristics. Its performance is investigated and compared with three existing harmony search algorithms. Experimental results reveal that proposed algorithm outperforms the existing approaches when applied to 15 benchmark functions. The effects of scalability, noise, and harmony memory size have also been investigated on four approaches of HS. The proposed algorithm is also employed for data clustering. Five real life datasets selected from UCI machine learning repository are used. The results show that, for data clustering, the proposed algorithm achieved results better than other algorithms.     

A genetic algorithm with local search strategy for improved detection of community structure

On the basis of modularity optimization, a genetic algorithm is proposed to detect community structure in networks by defining a local search operator. The local search operator emphasizes two features: one is that the connected nodes in a network should be located in the same community, while the other is “local selection” inspired by the mechanisms of efficient message delivery underlying the small‐world phenomenon. The results of community detection for some classic networks, such as Ucinet and Pajek networks, indicate that our algorithm achieves better community structure than other methodologies based on modularity optimization, such as the algorithms based on betweenness analysis, simulated annealing, or Tasgin and Bingol's genetic algorithm. © 2009 Wiley Periodicals, Inc. Complexity, 2010     

A secant algorithm with line search filter method for nonlinear optimization

Filter methods were initially designed for nonlinear programming problems by Fletcher and Leyffer. In this paper we propose a secant algorithm with line search filter method for nonlinear equality constrained optimization. The algorithm yields the global convergence under some reasonable conditions. By using the Lagrangian function value in the filter we establish that the proposed algorithm can overcome the Maratos effect without using second order correction step, so that fast local superlinear convergence to second order sufficient local solution is achieved. The primary numerical results are presented to confirm the robustness and efficiency of our approach.     

A line search trust-region algorithm with nonmonotone adaptive radius for a system of nonlinear equations

In this paper, a trust-region procedure is proposed for the solution of nonlinear equations. The proposed approach takes advantages of an effective adaptive trust-region radius and a nonmonotone strategy by combining both of them appropriately. It is believed that selecting an appropriate adaptive radius based on a suitable nonmonotone strategy can improve the efficiency and robustness of the trust-region frameworks as well as decrease the computational cost of the algorithm by decreasing the required number subproblems that must be solved. The global convergence and the local Q-quadratic convergence rate of the proposed approach are proved. Preliminary numerical results of the proposed algorithm are also reported which indicate the promising behavior of the new procedure for solving the nonlinear system.     

A reduced Hessian algorithm with line search filter method for nonlinear programming

This paper proposes a line search filter reduced Hessian method for nonlinear equality constrained optimization. The feature of the presented algorithm is that the reduced Hessian method is used to produce a search direction, a backtracking line search procedure to generate step size, some filtered rules to determine step acceptance, second order correction technique to reduce infeasibility and overcome the Maratos effects. It is shown that this algorithm does not suffer from the Maratos effects by using second order correction step, and under mild assumptions fast convergence to second order sufficient local solutions is achieved. The numerical experiment is reported to show the effectiveness of the proposed algorithm.