A memetic algorithm approach for solving the multidimensional multi-way number partitioning problem

In this paper, we describe a generalization of the multidimensional two-way number partitioning problem (MDTWNPP) where a set of vectors has to be partitioned into p sets (parts) such that the sums per every coordinate should be exactly or approximately equal. We will call this generalization the multidimensional multi-way number partitioning problem (MDMWNPP). Also, an efficient memetic algorithm (MA) heuristic is developed to solve the multidimensional multi-way number partitioning problem obtained by combining a genetic algorithm (GA) with a powerful local search (LS) procedure. The performances of our memetic algorithm have been compared with the existing numerical results obtained by CPLEX based on an integer linear programming formulation of the problem. The solution reveals that our proposed methodology performs very well in terms of both quality of the solutions obtained and the computational time compared with the previous method of solving the multidimensional two-way number partitioning problem.     

Heuristics and exact methods for number partitioning

Number partitioning is a classical NP-hard combinatorial optimization problem, whose solution is challenging for both exact and approximative methods. This work presents a new algorithm for number partitioning, based on ideas drawn from tree search, breadth first search, and beam search. A new set of benchmark instances for this problem is also proposed. The behavior of the new method on this and other testbeds is analyzed and compared to other well known heuristics and exact algorithms.     

Tabu search for graph partitioning

In this paper, we develop a tabu search procedure for solving the uniform graph partitioning problem. Tabu search, an abstract heuristic search method, has been shown to have promise in solving several NP-hard problems, such as job shop and flow shop scheduling, vehicle routing, quadratic assignment, and maximum satisfiability. We compare tabu search to other heuristic procedures for graph partitioning, and demonstrate that tabu search is superior to other solution approaches for the uniform graph partitioning problem both with respect to solution quality and computational requirements.     

Parallel local search

We present a survey of parallel local search algorithms in which we review the concepts that can be used to incorporate parallelism into local search. For this purpose we distinguish between single-walk and multiple-walk parallel local search and between asynchronous and synchronous parallelism. Within the class of single-walk algorithms we differentiate between multiple-step and single-step parallelism. To describe parallel local search we introduce the concepts of hyper neighborhood structures and distributed neighborhood structures. Furthermore, we present templates that capture most of the parallel local search algorithms proposed in the literature. Finally, we discuss some complexity issues related to parallel local search.     

A comparison of local search methods for flow shop scheduling

Local search techniques are widely used to obtain approximate solutions to a variety of combinatorial optimization problems. Two important categories of local search methods are neighbourhood search and genetic algorithms. Commonly used neighbourhood search methods include descent, threshold accepting, simulated annealing and tabu search. In this paper, we present a computational study that compares these four neighbourhood search methods, a genetic algorithm, and a hybrid method in which descent is incorporated into the genetic algorithm. The performance of these six local search methods is evaluated on the problem of scheduling jobs in a permutation flow shop to minimize the total weighted completion time. Based on the results of extensive computational tests, simulated annealing is found to generate better quality solutions than the other neighborhood search methods. However, the results also indicate that the hybrid genetic descent algorithm is superior to simulated annealing.     

Image registration with iterated local search

This contribution is devoted to the application of iterated local search to image registration, a very complex, real-world problem in the field of image processing. To do so, we first re-define this parameter estimation problem as a combinatorial optimization problem, then analyze the use of image-specific information to guide the search in the form of an heuristic function, and finally propose its solution by iterated local search. Our algorithm is tested by comparing its performance to that of two different baseline algorithms: iterative closest point, a well-known, image registration technique, a hybrid algorithm including the latter technique within a simulated annealing approach, a multi-start local search procedure, that allows us to check the influence of the search scheme considered in the problem solving, and a real coded genetic algorithm. Four different problem instances are tackled in the experimental study, resulting from two images and two transformations applied on them. Three parameter settings are analyzed in our approach in order to check three heuristic information scenarios where the heuristic is not used at all, is partially used or almost completely guides the search process, as well as two different number of iterations in the algorithms outer-inner loops. This work was partially supported by the Spanish Ministerio de Ciencia y Tecnología under project TIC2003-00877 (including FEDER fundings) and under Network HEUR TIC2002-10866-E.     

The linear ordering problem: Instances,search space analysis and algorithms

The linear ordering problem is an NP-hard problem that arises in a variety of applications. Due to its interest in practice, it has received considerable attention and a variety of algorithmic approaches to its solution have been proposed. In this paper we give a detailed search space analysis of available benchmark instance classes that have been used in various researches. The large fitness-distance correlations observed for many of these instances suggest that adaptive restart algorithms like iterated local search or memetic algorithms, which iteratively generate new starting solutions for a local search based on previous search experience, are promising candidates for obtaining high performing algorithms. We therefore experimentally compared two such algorithms and the final experimental results suggest that, in particular, the memetic algorithm is a new state-of-the-art approach to the linear ordering problem.     

Integrating local search techniques into mixed integer programming

We survey the main results of the authors PhD thesis that was supervised by Claude Le Pape (ILOG, France) and Philippe Michelon (Université dAvignon, France) and has been defended in June 2004. The dissertation is written in French and is available from the author. It introduces several strategies for integrating local search techniques into mixed integer programming, with an emphasis on generic algorithms.Received: June 2004, MSC classification: 90C11, 90C59     

带惩罚的相同容量k-均值问题的局部搜索算法

经典$k$-均值问题是一类应用广泛的聚类问题,它是指给定$\mathbb{R}^d$中观测点集合$D$和整数$k$,目的是在空间中寻找$k$个点作为中心集合$S$,使得集合$D$中的每个观测点到$S$中离它最近的中心的距离平方求和最小。这是个NP-难问题。经典$k$-均值问题有很多推广,本文研究的带惩罚的相同容量$k$-均值问题就是其中之一。与经典$k$-均值问题相比,惩罚性质是指每个观测点都给定惩罚费用,当某个观测点到最近中心的距离大于惩罚费用时,其对目标函数的贡献就用该观测点的惩罚费用来代替最近的距离的平方,相同容量约束要求每个中心至多连接$U$个观测点。针对这种问题,我们设计了局部搜索算法,该算法在至多选取$（3+\alpha）k$个中心的情况下,可以达到$\beta$-近似,其中,参数$\alpha>34$,$\beta>\frac{\alpha+34}{\alpha-34}$。     

带惩罚的相同容量k-均值问题的局部搜索算法

经典$k$-均值问题是一类应用广泛的聚类问题,它是指给定$\mathbb{R}^d$中观测点集合$D$和整数$k$,目的是在空间中寻找$k$个点作为中心集合$S$,使得集合$D$中的每个观测点到$S$中离它最近的中心的距离平方求和最小。这是个NP-难问题。经典$k$-均值问题有很多推广,本文研究的带惩罚的相同容量$k$-均值问题就是其中之一。与经典$k$-均值问题相比,惩罚性质是指每个观测点都给定惩罚费用,当某个观测点到最近中心的距离大于惩罚费用时,其对目标函数的贡献就用该观测点的惩罚费用来代替最近的距离的平方,相同容量约束要求每个中心至多连接$U$个观测点。针对这种问题,我们设计了局部搜索算法,该算法在至多选取$（3+\alpha）k$个中心的情况下,可以达到$\beta$-近似,其中,参数$\alpha>34$,$\beta>\frac{\alpha+34}{\alpha-34}$。     

Exploiting relaxation in local search for LABS

Branch-and-bound uses relaxation to prune search trees but sometimes scales poorly to large problems. Conversely, local search often scales well but may be unable to find optimal solutions. Both phenomena occur in the construction of low-autocorrelation binary sequences (LABS), a problem arising in communication engineering. This paper proposes a hybrid approach to optimization: using relaxation to prune local search spaces. An implementation gives very competitive results, showing the feasibility of the approach.     

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.     

An iterated local search algorithm for the vehicle routing problem with backhauls

The Vehicle Routing Problem with Backhauls (VRPB) is an extension of the VRP that deals with two types of customers: the consumers (linehaul) that request goods from the depot and the suppliers (backhaul) that send goods to the depot. In this paper, we propose a simple yet effective iterated local search algorithm for the VRPB. Its main component is an oscillating local search heuristic that has two main features. First, it explores a broad neighborhood structure at each iteration. This is efficiently done using a data structure that stores information about the set of neighboring solutions. Second, the heuristic performs constant transitions between feasible and infeasible portions of the solution space. These transitions are regulated by a dynamic adjustment of the penalty applied to infeasible solutions. An extensive statistical analysis was carried out in order to identify the most important components of the algorithm and to properly tune the values of their parameters. The results of the computational experiments carried out show that this algorithm is very competitive in comparison to the best metaheuristic algorithms for the VRPB. Additionally, new best solutions have been found for two instances in one of the benchmark sets. These results show that the performance of existing metaheuristic algorithms can be considerably improved by carrying out a thorough statistical analysis of their components. In particular, it shows that by expanding the exploration area and improving the efficiency of the local search heuristic, it is possible to develop simpler and faster metaheuristic algorithms without compromising the quality of the solutions obtained.     

The continuous reactive tabu search: Blending combinatorial optimization and stochastic search for global optimization

A novel algorithm for the global optimization of functions (C-RTS) is presented, in which a combinatorial optimization method cooperates with a stochastic local minimizer. The combinatorial optimization component, based on the Reactive Tabu Search recently proposed by the authors, locates the most promising boxes, in which starting points for the local minimizer are generated. In order to cover a wide spectrum of possible applications without user intervention, the method is designed with adaptive mechanisms: the box size is adapted to the local structure of the function to be optimized, the search parameters are adapted to obtain a proper balance of diversification and intensification. The algorithm is compared with some existing algorithms, and the experimental results are presented for a variety of benchmark tasks.     

Computational bounds for local search in combinatorial optimization

The results related to finding local optima in combinatorial optimization are overviewed. The class of polynomial-time local search problems (class PLS) is considered. By analogy with Cook’s theorem, the existence of most complicated problems in this class is established. The number of steps in local descent algorithms is estimated in the worst and average cases. The local search determination of exact and approximate solutions with guaranteed error bounds is discussed.     

Probabilistic diversification and intensification in local search for vehicle routing

This article presents a probabilistic technique to diversify, intensify, and parallelize a local search adapted for solving vehicle routing problems. This technique may be applied to a very wide variety of vehicle routing problems and local searches. It is shown that efficient first-level tabu searches for vehicle routing problems may be significantly improved with this technique. Moreover, the solutions produced by this technique may often be improved by a postoptimization technique presented in this article, too. The solutions of nearly forty problem instances of the literature have been improved.     

A guided local search metaheuristic for the team orienteering problem

In the team orienteering problem (TOP) a set of locations is given, each with a score. The goal is to determine a fixed number of routes, limited in length, that visit some locations and maximise the sum of the collected scores. This paper describes an algorithm that combines different local search heuristics to solve the TOP. Guided local search (GLS) is used to improve two of the proposed heuristics. An extra heuristic is added to regularly diversify the search in order to explore more areas of the solution space. The algorithm is compared with the best known heuristics of the literature and applied on a large problem set. The obtained results are almost of the same quality as the results of these heuristics but the computational time is reduced significantly. Applying GLS to solve the TOP appears to be a very promising technique. Furthermore, the usefulness of exploring more areas of the solution space is clearly demonstrated.     

From stars to comets: Improved local search for universal facility location

We improve the approximation ratio for the Universal Facility Location Problem to 6.702 by a local search algorithm with an extended pivot operation.     

Diversified local search strategy under scatter search framework for the probabilistic traveling salesman problem

This paper focuses on introducing a concept of diversified local search strategy under the scatter search framework for the probabilistic traveling salesman problem (PTSP). Different combinations of three commonly used local search methods in the PTSP, i.e., 1-shift, 2-opt, and 3-opt, were used to investigate its effects. A set of numerical experiments were conducted to test the validity of the proposed strategy based on randomly generated test instances. The numerical results and the permutation test showed that the diversified local search strategy, especially by combining 1-shift and 2-opt algorithms, can most effectively solve the homogeneous and heterogeneous PTSP in most of the tested instances in comparison with the single local search strategy under scatter search framework.