Search

In this paper, we study the quadratic assignment problem with a rank-one cost matrix (QAP-R1). Four integer-programming formulations are introduced of which three are assumed to have partial integer data. Unlike the standard quadratic assignment problem, some of our formulations can solve reasonably large instances of QAP-R1 with impressive running times and are faster than some metaheuristics. Pairwise relative strength of the LP relaxations of these formulations are also analyzed from theoretical and experimental points of view. Finally, we present a new metaheuristic algorithm to solve QAP-R1 along with its computational analysis. Our study offers the first systematic experimental analysis of integer-programming models and heuristics for QAP-R1. The benchmark instances with various characteristics generated for our study are made available to the public for future research work. Some new polynomially solvable special cases are also introduced. Summary of Contribution: This paper aims to advance our knowledge and ability in solving an important special case of the quadratic assignment problem. It shows how to exploit inherent properties of an optimization problem to achieve computational advantages, a strategy that was followed by researchers in model building and algorithm developments for decades. Our computational results attest to this time-tested general philosophy. The paper presents the first systematic computational study of the rank one quadratic assignment problem, along with new mathematical programming models and complexity analysis. We believe the theoretical and computational results of this paper will inspire further research on the topic and will be of significant value to practitioners using rank one quadratic assignment models.

The facility layout problem (FLP) is a very important class of NP-hard problems in operations research that deals with the optimal assignment of facilities to minimize transportation costs. The quadratic assignment problem (QAP) can model the FLP effectively. One of the FLPs is the hospital facility layout problem that aims to place comprehensive clinics, laboratories, and radiology units within predefined boundaries in a way that minimizes the cost of movement of patients and healthcare personnel. We are going to develop a hybrid method based on discrete differential evolution (DDE) algorithm for solving the QAP. In the existing DDE algorithms, certain issues such as premature convergence, stagnation, and exploitation mechanism have not been properly addressed. In this study, we first aim to discover the issues that make the current problem worse and to identify the best solution to the problem, and then we propose to develop a hybrid algorithm (HDDETS) by combining the DDE and tabu search (TS) algorithms to enhance the exploitation mechanism in the DDE algorithm. Then, the performance of the proposed HDDETS algorithm is evaluated by implementing on the benchmark instances from the QAPLIB website and by comparing with DDE and TS algorithms on the benchmark instances. It is found that the HDDETS algorithm has better performance than both the DDE and TS algorithms where the HDDETS has obtained 42 optimal and best-known solutions from 56 instances, while the DDE and TS algorithms have obtained 15 and 18 optimal and best-known solutions out of 56 instances, respectively. Finally, we propose to apply the proposed algorithm to find the optimal distributions of the advisory clinics inside the Azadi Hospital in Iraq that minimizes the total travel distance for patients when they move among these clinics. Our application shows that the proposed algorithm could find the best distribution of the hospital’s rooms, which are modeled as a QAP, with reduced total distance traveled by the patients.

The random walk’ correlation matrix of multi-objective combinatorial optimization problems utilizes both local structure and general statistics of the objective functions. Reckoning time of correlation, or the random walk of lag 0, is quadratic in problem size L and number of objectives D. The computational complexity of the correlation coefficients of mNK is $$O(D^2 K^2 L)$$ , and of mQAP is $$O(D^2 L^2)$$ , where K is the number of interacting bits. To compute the random walk of a lag larger than 0, we employ a weighted graph Laplacian that associates a mutation operator with the difference in the objective function. We calculate the expected objective vector of a neighbourhood function and the eigenvalues of the corresponding transition matrix. The computational complexity of random walk’s correlation coefficients is polynomial with the problem size L and the number of objectives D. The computational effort of the random walks correlation coefficients of mNK is $$O(2^K L D^2)$$ , whereas of mQAP is $$O(L^6 D^2)$$ . Numerical examples demonstrate the utilization of these techniques.

We consider new polynomially solvable cases of the well-known Quadratic Assignment Problem involving coefficient matrices with a special diagonal structure. By combining the new special cases with polynomially solvable special cases known in the literature we obtain a new and larger class of polynomially solvable special cases of the QAP where one of the two coefficient matrices involved is a Robinson matrix with an additional structural property: this matrix can be represented as a conic combination of cut matrices in a certain normal form. The other matrix is a conic combination of a monotone anti-Monge matrix and a down-benevolent Toeplitz matrix. We consider the recognition problem for the special class of Robinson matrices mentioned above and show that it can be solved in polynomial time.

The quadratic assignment problem (QAP) is a combinatorial optimization problem that arises in many real-world applications, such as equipment allocation in industry. The QAP is NP-hard and, in practice, one of the hardest combinatorial optimization problems to solve to optimality. Exact solutions of QAP are typically obtained by the branch-and-bound method. This method, however, potentially requires a high computational effort, and the use of good lower bounds is essential to prune the search tree. Branch-and-bound algorithms that use the dual-ascent procedure based on the level-2 reformulation linearization technique (RLT2) belong to the state of the art on exactly solving QAP. In this work, we propose a parallel implementation of that branch-and-bound algorithm. Our approach uses the Auction Algorithm of Bertsekas and Castañon to solve the linear assignment problems of RLT2, which allows us to take advantage of the massive parallel environment of graphics processing units to speed up the lower bound computation and implement some memory optimization techniques to address large-size problems. We report experimental results that show significant execution time reductions compared to previous works and allow us to provide, for the first time, exact solutions for two instances of QAP: tai35b and tai40b.

In this paper we study the complexity and domination analysis in the context of the \emph{bipartite quadratic assignment problem}. Two variants of the problem, denoted by BQAP1 and BQAP2, are investigated. A formula for calculating the average objective function value $\mathcal{A}$ of all solutions is presented whereas computing the median objective function value is shown to be NP-hard. We show that any heuristic algorithm that produces a solution with objective function value at most $\mathcal{A}$ has the domination ratio at least $\frac{1}{mn}$. Analogous results for the standard \emph{quadratic assignment problem} is an open question. We show that computing a solution whose objective function value is no worse than that of $n^mm^n-{\lceil\frac{n}{\alpha}\rceil}^{\lceil\frac{m}{\alpha}\rceil}{\lceil\frac{m}{\alpha}\rceil}^{\lceil\frac{n}{\alpha}\rceil}$ solutions of BQAP1 or $m^mn^n-{\lceil\frac{m}{\alpha}\rceil}^{\lceil\frac{m}{\alpha}\rceil}{\lceil\frac{n}{\alpha}\rceil}^{\lceil\frac{n}{\alpha}\rceil}$ solutions of BQAP2, is NP-hard for any fixed natural numbers $a$ and $b$ such that $\alpha=\frac{a}{b}>1$. However, a solution with the domination number $\Omega(m^{n-1}n^{m-1}+m^{n+1}n+mn^{m+1})$ for BQAP1 and $\Omega(m^{m-1}n^{n-1}+m^2n^{n}+m^mn^2)$ for BQAP2, can be found in $O(m^3n^3)$ time.

In this paper we propose a new lower bounding procedure for the Quadratic Assignment Problem based on a generalization of the well-known Gilomore-Lawler procedure for a higher order reformulation. Computational results on some benchmark instances show the strength of the new approach compared with other lower bounds.

We consider three known bounds for the quadratic assignment problem (QAP): an eigenvalue, a convex quadratic programming (CQP), and a semidefinite programming (SDP) bound. Since the last two bounds were not compared directly before, we prove that the SDP bound is stronger than the CQP bound. We then apply these to improve known bounds on a discrete energy minimization problem, reformulated as a QAP, which aims to minimize the potential energy between repulsive particles on a toric grid. Thus we are able to prove optimality for several configurations of particles and grid sizes, complementing earlier results by Bouman, Draisma and Van Leeuwaarden [ SIAM Journal on Discrete Mathematics, 27(3):1295--1312, 2013]. The semidefinite programs in question are too large to solve without pre-processing, and we use a symmetry reduction method by Parrilo and Permenter [Mathematical Programming, 181:51--84, 2020] to make computation of the SDP bounds possible.

A new repeated iterated tabu search for the quadratic assignment problem is developed. A comparative study of this algorithm and the best existing algorithms for solving this problem has shown its competitiveness with respect to both its performance and quality of solutions.

Definition of the equivalence between assignments and edit paths.Graph edit distance formulation as a quadratic assignment problem.New quadratic cost function for computing graph edit distance.Improvement of the accuracy of the approximation of graph edit distance.Approximation computable in reasonable time. The Graph Edit Distance (GED) is a flexible measure of dissimilarity between graphs which arises in error-correcting graph matching. It is defined from an optimal sequence of edit operations (edit path) transforming one graph into another. Unfortunately, the exact computation of this measure is NP-hard. In the last decade, several approaches were proposed to approximate the GED in polynomial time, mainly by solving linear programming problems. Among them, the bipartite GED received much attention. It is deduced from a linear sum assignment of the nodes of the two graphs, which can be efficiently computed by Hungarian-type algorithms. However, edit operations on nodes and edges are not handled simultaneously, which limits the accuracy of the approximation. To overcome this limitation, we propose to extend the linear assignment model to a quadratic one. This is achieved through the definition of a family of edit paths induced by assignments between nodes. We formally show that the GED, restricted to the paths in this family, is equivalent to a quadratic assignment problem. Since this problem is NP-hard, we propose to compute an approximate solution by adapting two algorithms: Integer Projected Fixed Point method and Graduated Non Convexity and Concavity Procedure. Experiments show that the proposed approach is generally able to reach a more accurate approximation of the exact GED than the bipartite GED, with a computational cost that is still affordable for graphs of non trivial sizes.

A new version of BLS algorithm is introduced.Levenshtein Distance (LD) metric is applied to the BLS heuristic.The QAP is optimized with several processors using OpenMP.Significant improvements are obtained.The BLS-OpenMP can be reported as one of the best heuristics for the QAP. The Quadratic Assignment Problem (QAP) is one of the most challenging NP-Hard combinatorial optimization problems. Circuit-layout design, transportation/traffic engineering, and assigning gates to airplanes are some of the interesting applications of the QAP. In this study, we introduce an enhanced version of a recent local search heuristic, Breakout Local Search Algorithm (BLS), by using the Levenshtein Distance metric for checking the similarity of the new starting points to previously explored QAP permutations. The similarity-checking process prevents the local search algorithm, BLS, from getting stuck in already-explored areas. In addition, the proposed BLS Algorithm (BLS-OpenMP) incorporates multi-threaded computation using OpenMP. The stagnation-aware search for the optimal solutions of the QAP is executed concurrently on several cores with diversified trajectories while considering their similarity to already-discovered local optima. The exploration of the search space is improved by selecting the starting points intelligently and speeding up the fitness evaluations linearly with number of processors/threads. BLS-OpenMP has been tested on the hardest 59 problem instances of the QAPLIB, and it obtained 57 of the best known results. The overall deviation of the achieved solutions from the best known results is 0.019% on average, which is a significant improvement compared with state-of-the-art algorithms.

The generalized quadratic assignment problem (GQAP) is the task of assigning a set of facilities to a set of locations such that the sum of the assignment and transportation costs is minimized. The facilities may have different space requirements, and the locations may have varying space capacities. Also, multiple facilities may be assigned to each location such that space capacity is not exceeded. In this paper, an application of the GQAP is presented for assigning a set of machines to a set of locations on the plant floor. Construction algorithms and a simple tabu search heuristic are developed for the GQAP. A set of test problems available in the literature was used to evaluate the performances of the TS heuristic using different construction algorithms. The results show that the simple TS heuristic is effective for solving the GQAP.

In this paper, we show that the quadratic assignment problem (QAP) can be reformulated to an equivalent rank constrained doubly nonnegative (DNN) problem. Under the framework of the difference of convex functions (DC) approach, a semi-proximal DC algorithm (DCA) is proposed for solving the relaxation of the rank constrained DNN problem whose subproblems can be solved by the semi-proximal augmented Lagrangian method (sPALM). We show that the generated sequence converges to a stationary point of the corresponding DC problem, which is feasible to the rank constrained DNN problem. Moreover, numerical experiments demonstrate that for most QAP instances, the proposed approach can find the global optimal solutions efficiently, and for others, the proposed algorithm is able to provide good feasible solutions in a reasonable time., Comment: 24 pages, 1 figure

The Quadratic Assignment Problem (QAP) is a well-known NP -hard combinatorial optimization problem that has received a lot of attention from the research community since it has many practical applications, such as allocation of facilities, design of electronic devices, etc. In this paper, we propose a hybrid approximate approach for the QAP based upon the framework of the Biased Random Key Genetic Algorithm. This hybrid approach includes an improvement method to be applied over the best individuals of the population in order to exploit the promising regions found in the search space. In the computational experiments, we evaluate the performance of our approach on widely known instances from the literature. In these experiments, we compare our approach against the best proposals from the related literature and we conclude that our approach is able to report high-quality solutions by means of short computational times.

We introduce a new binary quadratic program that subsumes the well known quadratic assignment problem. This problem is shown to be NP-hard when some associated parameters are restricted to simple values but solvable in polynomial time when some other parameter values are restricted. Three different neighborhood structures are introduced. A best solution in all these neighborhoods can be identified in polynomial time even though two of these neighborhoods are of exponential size. Different local search algorithms and their enhancements using tabu search are developed using these neighborhoods. Experimental analysis show that two hybrid algorithms obtained by combining these neighborhoods in specific ways yield results that are superior to the algorithms that use these neighborhoods separately. Extensive computational results and statistical analysis of the resulting data are presented.

Graphical abstractDisplay Omitted HighlightsA hybrid method combining great deluge and tabu search algorithms is proposed.This hybrid is supported with a two-stage external memory.First stage of acts as a short term memory that is frequently updated.Second stage acts as a long term memory that is updated less frequently.Elements of the second stage are maximally dissimilar in variable space. A two-stage memory architecture is maintained within the framework of great deluge algorithm for the solution of single-objective quadratic assignment problem. Search operators exploiting the accumulated experience in memory are also implemented to direct the search towards more promising regions of the solution space. The level-based acceptance criterion of the great deluge algorithm is applied for each best solution extracted in a particular iteration. The use of short- and long-term memory-based search supported by effective move operators resulted in a powerful combinatorial optimization algorithm. A successful variant of tabu search is employed as the local search method that is only applied over a few randomly selected memory elements when the second stage memory is updated. The success of the presented approach is illustrated using sets of well-known benchmark problems and evaluated in comparison to well-known combinatorial optimization algorithms. Experimental evaluations clearly demonstrate that the presented approach is a competitive and powerful alternative for solving quadratic assignment problems.

The convex and concave relaxations procedure (CCRP) was recently proposed to approximately solve the graph matching problem and exhibited a quite promising performance. To extend the CCRP to approximately solve the quadratic assignment problem (QAP), a major trouble is how to figure out the corresponding convex and concave relaxation functions. In this paper we will propose a general but simple QAP algorithm, and will then prove that the algorithm is exactly an type of CCRP algorithm, but without needing to figure out the convex or concave relaxation function in an explicit way. The proposed algorithm can be generally used on symmetric and asymmetric QAP's, and is simple to implement. Extensive experimental comparisons on the QAPLib benchmark data sets witness a state-of-the-art performance of the proposed algorithm.

International audience; Part-based image classification consists in representing categories by small sets of discriminative parts upon which a representation of the images is built. This paper addresses the question of how to automatically learn such parts from a set of labeled training images. We propose to cast the training of parts as a quadratic assignment problem in which optimal correspondences between image regions and parts are automatically learned. The paper analyses different assignment strategies and thoroughly evaluates them on two public datasets: Willow actions and MIT 67 scenes.

Teaching–Learning-Based Optimization (TLBO) is a novel swarm intelligence metaheuristic that is reported as an efficient solution method for many optimization problems. It consists of two phases where all individuals are trained by a teacher in the first phase and interact with classmates to improve their knowledge level in the second phase. In this study, we propose a set of TLBO-based hybrid algorithms to solve the challenging combinatorial optimization problem, Quadratic Assignment. Individuals are trained with recombination operators and later a Robust Tabu Search engine processes them. The performances of sequential and parallel TLBO-based hybrid algorithms are compared with those of state-of-the-art metaheuristics in terms of the best solution and computational effort. It is shown experimentally that the performance of the proposed algorithms are competitive with the best reported algorithms for the solution of the Quadratic Assignment Problem with which many real life problems can be modeled.

We propose an exact algorithm to find the optimal solution to the grey pattern problem which is a special instance of the quadratic assignment problem. A very effective branch and bound algorithm is constructed. The special structure of the problem is exploited and optimal solutions for many problem instances established. One instance based on 64 facilities, organized in an 8 by 8 pattern, that was extensively researched in the literature was optimally solved in 15 s of computer time while existing papers report solution times of hours.

Instead of using traditional (two-parent) crossover operator, multi-parent crossover operator is used in genetic algorithms to improve solution quality for many numerical optimization problems. However, very few literatures are available on multi-parent crossover operator for combinatorial optimization problems, especially, quadratic assignment problem (QAP). This paper proposes a multi-parent extension of the sequential constructive crossover (MPSCX), which is a generalization of the traditional sequential constructive crossover (SCX), for the QAP. Then a multi-parent genetic algorithm (MPGA) using MPSCX is developed. Experimental results on ten QAPLIB instances show that our MPGA significantly improves GA using SCX by up to 1.75 % in average assignment cost with maximum of 2.79 % away from the best known solution value. Finally, the efficiency of our MPGA is compared against MPGA using an existing multi-parent crossover for the problem. Experimental results show that our MPGA is better.

In genetic algorithms crossover is the most important operator where pair of chromosomes and crossover site along their common length are selected randomly. Then the information after the crossover site of the parent chromosomes is swapped. On the other hand, mutation operator randomly alters some genes of a chromosome, and thus diversifies the search space. We consider three crossover and ten mutation operators for the genetic algorithms which are then compared for the quadratic assignment problem on some benchmark QAPLIB instances. The experimental study shows the effectiveness of the sequential constructive crossover and the adaptive mutation operators for the problem.

The quadratic assignment problem is an NP-hard discrete optimization program that has been extensively studied for over 50?years. It has a variety of applications in many fields, but has proven itself extremely challenging to solve. As a result, an area of research has been to identify special cases which admit efficient solution strategies. This paper examines four such cases, and shows how each can be explained in terms of the dual region to the continuous relaxation of a classical linear reformulation of the problem known as the level-1 RLT representation. The explanations allow for simplifications and/or generalizations of the conditions defining the special cases.

We propose two variations on particle swarm optimization (PSO): the use of a heuristic function as an additional biasing term in PSO solution construction; and the use of a local search step in the PSO algorithm. We apply these variations to the hierarchical PSO model and evaluate them on the quadratic assignment problem (QAP). We compare the performance of our method to diversified-restart robust tabu search (DivTS), one of the leading approaches at present for the QAP. Our experimental results, using instances from the QAPLIB instance library, indicate that our approach performs competitively with DivTS.

The generalized quadratic assignment problem (GQAP) concerns with assigning m facilities to n locations. Multiple facilities can be assigned to a single location as long as the space allows. The aim is to find a way of assignment such that the total cost involved is minimized while respecting the space constraints. The total cost composes of installation and transportation costs, which is obtained by multiplying the facility flows and distances. GQAP is the generalization of the well-known quadratic assignment problem, which requires the number of facilities and locations to be the same and exactly one facility is assigned to one location. This paper presents a hybrid metaheuristic for solving the GQAP. The proposed algorithm hybridizes biogeography-based optimization with tabu search. Experimental results show that the hybrid metaheuristic is able to solve problem instances obtained from a manufacturing company, within reasonable computational times.

International audience; The Quadratic Assignment Problem is at the core of several real-life applications. Finding an optimal assignment is computationally very difficult, for many useful instances. The best results are obtained with hybrid heuristics, which result in complex solvers. We propose an alternate solution where hybridization is obtain by means of parallelism and cooperation between simple single-heuristic solvers. We present experimental evidence that this approach is very efficient and can effectively solve a wide variety of hard problems, often surpassing state-of-the-art systems.

International audience; Several real-life applications can be stated in terms of the Quadratic Assignment Problem. Finding an optimal assignment is com-putationally very difficult, for many useful instances. We address this problem using a local search technique, based on Extremal Optimization and present experimental evidence that this approach is competitive. Moreover, cooperative parallel versions of our solver improve performance so much that large and hard instances can be solved quickly.

The quadratic assignment problem (QAP) is one of the hardest combinatorial optimization problems. Its range of applications is wide, including facility location, keyboard layout, and various other domains. The key success factor of specialized branch-and-bound frameworks for minimizing QAPs is an efficient implementation of a strong lower bound. In this paper, we propose a lower-bound-preserving transformation of a QAP to a different quadratic problem that allows for small and efficiently solvable SDP relaxations. This transformation is self-tightening in a branch-and-bound process.

Nowadays, the quadratic assignment problem QAP is widely considered as one of the hardest of the NP-hard problems. One of the main reasons for this consideration can be found in the enormous difficulty of computing good quality bounds for branch-and-bound algorithms. The practice shows that even with the power of modern computers QAPs of size $$n > 30$$ are typically recognized as huge computational problems. In this work, we are concerned with the design of a new low-dimensional semidefinite programming relaxation for the computation of lower bounds of the QAP. We discuss ways to improve the bounding program upon its semidefinite relaxation base and give numerical examples to demonstrate its applicability.

Quadratic Assignment Problem (QAP) is a discrete search space problem where one has to find the perfect feasible permutation of locations which minimizes the objective function. This paper solves the quadratic assignment problem using bat algorithm. Bat algorithm cannot be directly applied for solving QAP as it generates a solution for continuous search space problems. Therefore, we use a heuristic rule called Smallest Position Value (SPV) to enable our algorithm for finding solutions for sequencing problems. The modified algorithm completely outperforms the particle swarm optimization algorithm in all the cases considered.

International audience; We consider the (QAP) that consists in minimizing a quadratic function subject to assignment constraints where the variables are binary. In this paper, we build two families of equivalent quadratic convex formulations of (QAP). The continuous relaxation of each equivalent formulation is then a convex problem and can be used within a B&B. In this work, we focus on finding the “best” equivalent formulation within each family, and we prove that it can be computed using semidefinite programming. Finally, we get two convex formulations of (QAP) that differ from their sizes and from the tightness of their continuous relaxation bound. We present computational experiments that prove the practical usefulness of using quadratic convex formulation to solve instances of (QAP) of medium sizes.

The quadratic assignment problem (QAP) is a NP-hard combinatorial optimization problem. Genetic algorithm (GA) is one of the best algorithms to deal with such difficult problems. This paper presents an improved GA for finding effective solution to the QAP. As starting with a good initial population leads faster convergence of GA, we use sequential sampling algorithm for generating initial population. In GA, crossover operator plays very important role and sequential constructive crossover (SCX) is found to be one of the best crossover operators for solving the QAP. We propose a restricted improvement of the SCX using a combined mutation operator. Also, an adaptive mutation operator is proposed to diversify the search space intelligently. Experimental results on some benchmark QAPLIB instances show the effectiveness of the improved algorithm.

Recent progress in solving quadratic assignment problems (QAPs) from the QAPLIB (Quadratic Assignment Problem Library) test set has come from mixed-integer linear or quadratic programming models that are solved in a branch-and-bound framework. Semidefinite programming (SDP) bounds for QAPs have also been studied in some detail, but their computational impact has been limited so far, mostly because of the restrictive size of the early relaxations. Some recent progress has been made by studying smaller SDP relaxations and by exploiting group symmetry in the QAP data. In this work, we introduce a new SDP relaxation, where the matrix variables are only of the order of the QAP dimension, and we show how one may exploit group symmetry in the problem data for this relaxation. We also provide a detailed numerical comparison with related bounds from the literature. In particular, we compute the best-known lower bounds for two QAPLIB instances.

Recently, distance-based exponential probability models have demonstrated their validity in the context of estimation of distribution algorithms when solving permutationbased combinatorial optimisation problems. However, despite their successful performance, some of these models are unimodal, and, therefore, they might not be flexible enough to model the different modalities that may be represented in heterogeneous populations. In this paper, we address the particular case of the Generalized Mallows models under the Cayley distance, and propose mixtures of these models in the context of estimation of distribution algorithms. In order to evaluate their competitiveness, we considered the quadratic assignment problem as a case of study, and conducted experiments over a set of 90 instances for four different configurations of mixtures. Results reveal that the EDA with mixtures is able to outperform the Generalized Mallows EDA, especially in large instances.

Previous work introduced an approach called TBHPSO, which combines Hierarchical Particle Swarm Optimization (HPSO) with Tabu Local Search and a heuristic bias term, to the Quadratic Assignment Problem (QAP). Specifically, in TBHPSO, a Robust Tabu Local Search is applied to the top particle in the hierarchy in each PSO iteration; in addition, a heuristic “goodness” function, similar to that used in Ant Colony Optimization, is used to bias the PSO velocity update equation. In previous work, TBHPSO was found to perform significantly better than Diversified-Restart Robust Tabu Search (DivTS), a state-of-the-art technique for the QAP. In this paper, we introduce three variations to TBHPSO. The first variation (RTBHPSO) aims to increase search diversity by applying Tabu Local Search to a randomly chosen particle rather than always applying it to the root of the HPSO hierarchy. The second variation (DTBHPSO) applies DivTS (instead of RTS) to the top particle in the hierarchy, while the third variation first selects a random particle and then probabilistically selects either DivTS or RTS to apply to it. The performance of our proposed variations is compared against the original TBHPSO, keeping the CPU time fixed for both methods in each comparison, using 31 problem instances from the QAPLib instance library.

The Quadratic Assignment Problem (QAP) is a well-studied, NP-Hard combinatorial optimization problem with practical applications in timetabling, scheduling, logistics, circuit design and data visualisation, to name a few. In this paper a Memetic Algorithm is described, which utilises a ternary tree structure for its population and uses a Tabu Search as its local improvement strategy. The Tabu Search is also run in parallel, significantly reducing the running time of the algorithm. The ternary tree not only stores the individuals within the population, but the inherent structure within this tree also determines parent selection for crossover. A small number of rules, which include fitness and diversity-based rules, govern whether a newly produced solution remains within the population, or whether it is discarded. These key features are tested against a basic Memetic Algorithm using the instances from the QAP library, QAPLIB, and have shown to significantly improve the performance in terms of both time and solution quality. The best version of the Memetic Algorithm is shown to perform competitively with some of the state-of-the-art algorithms for the QAP from the literature, with grid-based and real-life instances shown to be solved very efficiently and effectively by the presented algorithms.

In the last decade, many works in combinatorial optimisation have shown that, due to the advances in multi-objective optimisation, the algorithms in this field could be used for solving single-objective problems. In this sense, a number of papers have proposed multi-objectivising single-objective problems in order to apply multi-objectivisation schemes in their optimisation. In this paper, we follow this idea by presenting a method to multi-objectivise single-objective problems based on an elementary landscape decomposition of their objective function. In order to illustrate this procedure, we consider the elementary landscape decomposition of the Quadratic Assignment Problem under the interchange neighbourhood. In particular, we propose reformulating the QAP as a multi-objective problem, where each elementary landscape in the decomposition is an independent function to be optimised. In order to validate this multi-objectivisation scheme, we implement a version of NSGA-II for solving the multi-objective QAP, and compare its performance with that of a GA on the single-objective QAP. Conducted experiments show that the multi-objective approach is better than the single-objective approach for some types of instances.

This chapter has three objectives: (1) to describe how social network analyses (SNA) can be used to explore complexity dynamics in education; (2) to provide a primer on SNA methods; and (3) to explore statistical procedures for hypothesis testing with SNA. SNA has experienced increasing popularity in recent years, but resources available to researchers wanting to learn about this methodology are sparse. That which is available typically fails to link SNA to complexity theory, although this would seem an obvious context. This chapter briefly describes major principles of complexity theory and how network analyses are useful for exploring social dynamics. We then explain what SNA is, the types of analyses it performs, and its various uses. This section delves into issues such as designing SNA analyses, data collection procedures, and converting non-matrix data for use in SNA. Lastly, the chapter describes statistical procedures for analyzing network data. In particular, we explain how to conduct multiple regression quadratic assignment procedures and p* to test hypotheses about network dynamics. Issues of using network coefficients with traditional, variable-based statistics are discussed. Examples of applicable research questions and research studies are provided to help readers formulate questions and research designs.

Optimising in many-objective search spaces, i.e. search spaces with more than three objectives, is a challenging task. Scalarization functions transform the multi-objective search space into a single objective search space. In order to scale-up optimisation in many-objective search spaces, we use Cartesian product of scalarization functions or simpler product functions to reduce the number of objectives of the search space. Stochastic product local search (SprLS) uses product functions to evaluate solutions within a local search run with the goal of generating the entire Pareto front. To improve the performance of SprLS algorithms: 1) we pursuit a fixed set the product function that improves the most the performance of the algorithm, or 2) we adapt the direction of the component scalarization functions using solutions in the current (possible suboptimal) Pareto front. We compare the performance of local search algorithms on many-objective quadratic assignment instances with correlated flow matrices.

This paper studies the cross-correlation of two real-valued functions whose common domain is a set on which a finite group acts transitively. Such a group action generalizes, e.g., circular time shifts for discrete periodic time signals or spatial translations in the context of image processing. Cross-correlations can be reformulated as convolutions in group algebras and are thus transformable into the spectral domain via Fast Fourier Transforms. The resulting general discrete cross-correlation theorem will serve as the basis for a unified FFT-based approach to maximum assignment problems related to transitive finite group actions. Our main focus is on those assignment problems arising from the actions of symmetric groups on the cosets of Young subgroups. The “simplest” nontrivial representatives of the latter class of problems are the NP-hard symmetric and asymmetric maximum quadratic assignment problem. We present a systematic spectral approach in terms of the representation theory of symmetric groups to solve such assignment problems. This generalizes and algorithmically improves Kondor's spectral branch-and-bound approach (SODA, 2010) to the exact solution of the asymmetric maximum quadratic assignment problem.

Το Shortest Superstring Problem (SSP) είναι ένα πρόβλημα συνδυαστικής βελτιστοποίησης που έχει προσελκύσει το ενδιαφέρων πολλών ερευνητών, λόγω των εφαρμογών του. Μπορεί να χρησιμοποιηθεί σε προβλήματα Υπολογιστικής Μοριακής Βιολογίας όπως η αλληλούχιση του DNA και σε προβλήματα της επιστήμης υπολογιστών όπως η συμπίεση δεδομένων. Το SSP είναι ένα NP-hard πρόβλημα. Ένα άρθρο ανασκόπησης για το SSP παρουσιάζεται στο πρώτο κεφάλαιο της παρούσας διατριβής με έναν περιεκτικό και σαφή τρόπο, καλύπτοντας ολόκληρη τη σχετική βιβλιογραφία, αναδεικνύοντας την κατακτημένη γνώση και βοηθώντας στην μελλοντική έρευνα.Η μέθοδος GRASP (Greedy Randomized Adaptive Search Procedure) είναι μια επαναληπτική ευρετική μέθοδος για συνδυαστική βελτιστοποίηση. Η μέθοδος Path Relinking (PR) αποτελεί έναν τρόπο ενοποίησης των στρατηγικών εντατικοποίησης και διαφοροποίησης στην αναζήτηση για βέλτιστες λύσεις. Η PR στα πλαίσια του GRASP εισήχθη ως μηχανισμός μνήμης για την αξιοποίηση των δεδομένων από καλές λύσεις που έχουν ήδη βρεθεί. Στο δεύτερο κεφάλαιο, παρουσιάζεται η υλοποίηση της μεθόδου GRASP με PR για το SSP. Η νέα μέθοδος λύνει στιγμιότυπα μεγάλης κλίμακας και υπερτερεί του φυσικού άπληστου αλγόριθμου στη συντριπτική πλειοψηφία των στιγμιοτύπων που δοκιμάστηκαν. Η προτεινόμενη μέθοδος είναι ικανή να παράγει πολλαπλές λύσεις κοντά στο βέλτιστο, γεγονός το οποίο είναι σημαντικό για την πρακτική της αλληλούχισης του DNA και επιτρέπει μια φυσική και εύκολη παράλληλη υλοποίηση. Ένα σύνολο αναφοράς στιγμιοτύπων με γνωστή βέλτιστη λύση κατασκευάστηκε χρησιμοποιώντας μια νέα Διατύπωση Ακέραιου Προγραμματισμού (Integer Programming Formulation) για το SSP.Η οικογένεια των γράφων επικάλυψης αποτελεί ένα κατάλληλο είδος δομής δεδομένων για την περίπτωση του SSP. Έχουν εφαρμογές στην αλληλούχιση γονιδιώματος, στην συμπίεση ακολουθιών και στον χρονοπρογραμματισμό μηχανών. Ένας κατευθυνόμενος γράφος με βάρη είναι γράφος επικάλυψης αν υπάρχει ένα σύνολο από ακολουθίες, οι οποίες βρίσκονται σε ένα προς ένα αντιστοιχία με τις κορυφές του γράφου, έτσι ώστε κάθε βάρος του γράφου να ισούται με την επικάλυψη μεταξύ των αντίστοιχων ακολουθιών. Στο τρίτο κεφάλαιο της παρούσας διατριβής, παρουσιάζεται ένα θεώρημα χαρακτηρισμού των γράφων επικάλυψης και ο αντίστοιχος αλγόριθμος αναγνώρισής τους.Το Quadratic Assignment Problem (QAP) είναι ένα από τα δυσκολότερα προβλήματα συνδυαστικής βελτιστοποίησης. Το QAP είναι ένα NP-hard πρόβλημα, ενώ η εύρεση ενός ε-προσεγγιστικού αλγόριθμου για αυτό είναι επίσης δύσκολη. Ο κλασικός άπληστος αλγόριθμος για διακριτά προβλήματα βελτιστοποίησης όπου η βέλτιστη λύση είναι ένα μεγιστοτικό ανεξάρτητο υποσύνολο ενός πεπερασμένου συνόλου βάσης με στοιχεία με βάρη, μπορεί να οριστεί με δύο διαφορετικούς τρόπους που είναι δυϊκοί ο ένας προς το άλλο. Τον άπληστο-εισαγωγής (greedy-in) αλγόριθμο, όπου μια λύση κατασκευάζεται από ένα κενό σύνολο με την εισαγωγή του επόμενου καλύτερου στοιχείου, ενός κάθε φορά, μέχρι να προκύψει μια μη εφικτή λύση και τον άπληστο-εξαγωγής (greedy-out) αλγόριθμο, όπου ξεκινώντας από το σύνολο βάσης, διαγράφεται το επόμενο χειρότερο στοιχείο, ένα κάθε φορά, μέχρι να προκύψει κάποια εφικτή λύση. Έχει αποδειχτεί ότι ενώ ο πρώτος αλγόριθμος παρέχει έναν παράγοντα προσέγγισης για τα προβλήματα μεγιστοποίησης, η απόδοσή του στην χειρότερη περίπτωση δεν είναι φραγμένη για τα προβλήματα ελαχιστοποίησης και το αντίστροφο για τον δεύτερο αλγόριθμο. Στο τέταρτο κεφάλαιο αυτής της διατριβής, παρουσιάζεται ο άπληστος-εξαγωγής αλγόριθμος για το QAP, αφότου αναπτύσσεται ένας συνδυαστικός χαρακτηρισμός των λύσεων του προβλήματος.

We introduce and study the combinatorial optimization problem with interaction costs (COPIC). COPIC is the problem of finding two combinatorial structures, one from each of two given families, such that the sum of their independent linear costs and the interaction costs between elements of the two selected structures is minimized. COPIC generalizes the quadratic assignment problem and many other well studied combinatorial optimization problems, and hence covers many real world applications. We show how various topics from different areas in the literature can be formulated as special cases of COPIC. The main contributions of this paper are results on the computational complexity and approximability of COPIC for different families of combinatorial structures (e.g. spanning trees, paths, matroids), and special structures of the interaction costs. More specifically, we analyze the complexity if the interaction cost matrix is parameterized by its rank and if it is a diagonal matrix. Also, we determine the structure of the intersection cost matrix, such that COPIC is equivalent to independently solving linear optimization problems for the two given families of combinatorial structures.

The Reformulation Linearization Technique (RLT) of Sherali and Adams (Manag Sci 32(10):1274–1290, 1986; SIAM J Discrete Math 3(3):411–430, 1990), when applied to a pure 0–1 quadratic optimization problem with linear constraints (P), constructs a hierarchy of LP (i.e., continuous and linear) models of increasing sizes. These provide monotonically improving continuous bounds on the optimal value of (P) as the level, i.e., the stage in the process, increases. When the level reaches the dimension of the original solution space, the last model provides an LP bound equal to the IP optimum. In practice, unfortunately, the problem size increases so rapidly that for large instances, even computing bounds for RLT models of level k (called RLTk) for small k may be challenging. Their size and their complexity increase drastically with k. To our knowledge, only results for bounds of levels 1, 2, and 3 have been reported in the literature. We are proposing, for certain quadratic problem types, a way of producing stronger bounds than continuous RLT1 bounds in a fraction of the time it would take to compute continuous RLT2 bounds. The approach consists in applying a specific decomposable Lagrangean relaxation to a specially constructed RLT1-type linear 0–1 model. If the overall Lagrangean problem does not have the integrality property, and if it can be solved as a 0–1 rather than a continuous problem, one may be able to obtain 0–1 RLT1 bounds of roughly the same quality as standard continuous RLT2 bounds, but in a fraction of the time and with much smaller storage requirements. If one actually decomposes the Lagrangean relaxation model, this two-step procedure, reformulation plus decomposed Lagrangean relaxation, will produce linear 0–1 Lagrangean subproblems with a dimension no larger than that of the original model. We first present numerical results for the Crossdock Door Assignment Problem, a special case of the Generalized Quadratic Assignment Problem. These show that just solving one Lagrangean relaxation problem in 0–1 variables produces a bound close to or better than the standard continuous RLT2 bound (when available) but much faster, especially for larger instances, even if one does not actually decompose the Lagrangean problem. We then present numerical results for the 0–1 quadratic knapsack problem, for which no RLT2 bounds are available to our knowledge, but we show that solving an initial Lagrangean relaxation of a specific 0–1 RLT1 decomposable model drastically improves the quality of the bounds. In both cases, solving the fully decomposed rather than the decomposable Lagrangean problem to optimality will make it feasible to compute such bounds for instances much too large for computing the standard continuous RLT2 bounds.