Variable Search Research Articles

Effective and Efficient Algorithms for Flexible Aggregate Similarity Search in High Dimensional Spaces

Numerous applications in different fields, such as spatial databases, multimedia databases, data mining, and recommender systems, may benefit from efficient and effective aggregate similarity search, also known as aggregate nearest neighbor (AggNN) search. Given a group of query objects $Q$ , the goal of AggNN is to retrieve the $k$ most similar objects from the database, where the underlying similarity measure is defined as an aggregation (usually sum or max ) of the distances between the retrieved objects and every query object in $Q$ . Recently, the problem was generalized so as to retrieve the $k$ objects which are most similar to a fixed proportion of the elements of $Q$ . This variant of aggregate similarity search is referred to as “flexible AggNN”, or FANN. In this work, we propose two approximation algorithms, one for the sum variant of FANN, and the other for the max variant. Extensive experiments are provided showing that, relative to state-of-the-art approaches (both exact and approximate), our algorithms produce query results with good accuracy, while at the same time being very efficient.

An effective new island model genetic algorithm for job shop scheduling problem

This paper presents an effective new island model genetic algorithm to solve the well-known job shop scheduling problem with the objective of minimizing the makespan. To improve the effectiveness of the classical island model genetic algorithm, we have proposed a new naturally inspired evolution model and a new naturally inspired migration selection mechanism that are capable of improving the search diversification and delaying the premature convergence. In the proposed evolution model, islands employ different evolution methods during their self-adaptation phases, rather than employing the same methods. In the proposed migration selection mechanism, worst individuals who are least adapted to their environments migrate first, hoping in finding a better chance to live in a more suitable environment that imposes a more suitable self-adaptation method on them. The proposed algorithm is tested on 52 benchmark instances, with the proposed evolution model and migration selection mechanism, and without them using the classical alternatives, and also compared with other algorithms recently reported in the literature. Computational results verify the improvements achieved by the proposed evolution model and migration selection mechanism, and show the superiority of the proposed algorithm over the others in terms of effectiveness.

An improved harmony search algorithm with dynamically varying bandwidth

The present work demonstrates a new variant of the harmony search (HS) algorithm where bandwidth (BW) is one of the deciding factors for the time complexity and the performance of the algorithm. The BW needs to have both explorative and exploitative characteristics. The ideology is to use a large BW to search in the full domain and to adjust the BW dynamically closer to the optimal solution. After trying a series of approaches, a methodology inspired by the functioning of a low-pass filter showed satisfactory results. This approach was implemented in the self-adaptive improved harmony search (SIHS) algorithm and tested on several benchmark functions. Compared to the existing HS algorithm and its variants, SIHS showed better performance on most of the test functions. Thereafter, the algorithm was applied to geometric parameter optimization of a friction stir welding tool.

A multi-agent based optimization method applied to the quadratic assignment problem

Inspired by the idea of interacting intelligent agents of a multi-agent system, we introduce a multi-agent based optimization method applied to the quadratic assignment problem (MAOM-QAP). MAOM-QAP is composed of several agents (decision-maker agent, local search agents, crossover agents and perturbation agent) which are designed for the purpose of intensified and diversified search activities. With the help of a reinforcement learning mechanism, MAOM-QAP dynamically decides the most suitable agent to activate according to the state of search process. Under the coordination of the decision-maker agent, the other agents fulfill dedicated search tasks. The performance of the proposed approach is assessed on the set of well-known QAP benchmark instances, and compared with the most advanced QAP methods of the literature. The ideas proposed in this work are rather general and could be adapted to other optimization tasks. This work opens the way for designing new distributed intelligent systems for tackling other complex search problems.

Multilayer variable neighborhood search for two‐level uncapacitated facility location problems with single assignment

We develop a variant of the variable neighborhood search (VNS) metaheuristic called the multilayer VNS (MLVNS). It consists in partitioning the neighborhood structures into multiple layers. For each layer , a VNS defined on the associated neighborhood structures is invoked, each move being evaluated and completed by a recursive call to the MLVNS at layer . A specific MLVNS is developed to solve approximately a class of two‐level uncapacitated facility location problems with single assignment (TUFLPS), when only mild assumptions are imposed on the cost functions. Two special cases are used to illustrate the efficiency of the MLVNS: the classical TUFLPS and a problem with modular costs derived from a real‐life case. To assess the efficiency of the MLVNS, computational results on a large set of instances are compared with those obtained by slope scaling heuristic methods and by solving integer programming models using a state‐of‐the‐art commercial solver. © 2015 Wiley Periodicals, Inc. NETWORKS, Vol. 66(3), 214–234 2015

A great deluge and tabu search hybrid with two-stage memory support for quadratic assignment problem

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.

Heuristic approaches for biobjective mixed 0–1 integer linear programming problems

In this study, biobjective mixed 0–1 integer linear programming problems are considered and two heuristic approaches are presented to find the Pareto frontier of these problems. The first heuristic is a variant of the variable neighborhood search and explores the k-neighbors of a feasible solution (in terms of binary variables) to find the extreme supported Pareto points. The second heuristic is adapted from the local branching method, which is well-known in single objective mixed 0–1 integer linear programming. Finally, an algorithm is proposed to find Pareto segments of outcome line segments of these heuristics. A computational analysis is performed by using some test problems from the literature and the results are presented.

Exploiting visual saliency for increasing diversity of image retrieval results

Diversification of search results allows for better and faster search, gaining knowledge about different perspectives and viewpoints on retrieved information sources. Recently various methods for diversification of image retrieval results have been proposed, mainly using textual information or techniques imported from the natural language processing domain. However, images contain much more information than their textual descriptions and the use of visual features deserves special attention in this context. Visual saliency provides information about parts of the image perceived as most important, which are instinctively targeted by humans when shooting a photo or looking at a picture. For this reason we propose to exploit such information to improve diversification of search results. To this purpose, we introduce a saliency-based method to re-rank the results of a query and we show that it can achieve significantly better performances as compared to the baseline approach. Experimental validation conducted on a number of queries applied to various datasets demonstrates the potential of the use of saliency information for the diversification of image retrieval results.

Context-Based Diversification for Keyword Queries Over XML Data

While keyword query empowers ordinary users to search vast amount of data, the ambiguity of keyword query makes it difficult to effectively answer keyword queries, especially for short and vague keyword queries. To address this challenging problem, in this paper we propose an approach that automatically diversifies XML keyword search based on its different contexts in the XML data. Given a short and vague keyword query and XML data to be searched, we first derive keyword search candidates of the query by a simple feature selection model. And then, we design an effective XML keyword search diversification model to measure the quality of each candidate. After that, two efficient algorithms are proposed to incrementally compute top-k qualified query candidates as the diversified search intentions. Two selection criteria are targeted: the k selected query candidates are most relevant to the given query while they have to cover maximal number of distinct results. At last, a comprehensive evaluation on real and synthetic data sets demonstrates the effectiveness of our proposed diversification model and the efficiency of our algorithms.

An adaptive variable neighborhood search for solving the multi-objective node placement problem

In this paper, we address the multi-objective node placement (MONP) problem that consists in extending an existing heterogeneous network while optimizing three conflicting objectives. Numerous devices' types are to be deployed in an area of interest in order to ensure the network connectivity and to satisfy users' demands. As the MONP problem is NP-Hard, we adapt a new variant of the multiobjective variable neighborhood search (MO-VNS) algorithm. The main idea is to automatically choose the most promising neighborhood structure by associating a rating to each one. This value is dynamically updated according to the number of so far generated non-dominated solutions. The empirical validation is done using a simulation environment called Inform Lab. A comparison to an existing multi-objective genetic approach is performed based on real instances of maritime surveillance application. The experimental results show that both methods have comparable performances based the unary hyper-volume metric (IH−). A slight improvement is noticed for the (MO-VNS) regarding medium sized instances.

Development of a new approach, evolutionary harmony search algorithm, for the LPO problem

This paper proposes a new variant of harmony search (HS) algorithm i.e. evolutionary harmony search (EHS), for exploiting in the loading pattern optimization (LPO) problem. The main innovations of EHS are the consideration of current best harmony vector in creating the new solution vector in any iteration and applying a pitch adjustment approach using a mutation strategy borrowed from the realm of the differential evolution (DE) algorithms, both with dynamic probabilities. Reactor core pattern optimization has been done using EHS for two test cases including KWU PWR and VVER-440. In order to represent the EHS capability to gain improved results in the LPO problem, a comparison is performed between results of EHS and a recent developed HS algorithm i.e. self-adaptive global best harmony search (SGHS). Numerical results show a major and distinctive enhancement of the proposed approach, EHS, in obtaining convergent results in comparison to SGHS approach. As a result, I can recommend the checking of EHS performance in other optimization problems.

A breakout local search (BLS) method for solving the assembly sequence planning problem

Being one of the main subproblems of the broader Assembly Planning (AP) problem, Assembly Sequence Planning (ASP) is defined as the process of computing a sequence of assembly motions for constituent parts of an assembled final product. ASP is proven to be NP-complete and thus its effective and efficient solution has been a challenge for the researchers in the field. However, despite the existence of numerous works on solving the ASP, the topology, structure, and complexity of the problem׳s search space (i.e., the fitness landscape) has not been studied yet. In this article, the fitness landscape of the ASP problem is analyzed for five typical assembled products through various distribution and correlation statistical measures, which reveals that locally optimal assembly sequences are distributed in the problem׳s landscape nearly uniformly. Based on this result, a suitable optimization algorithm called Breakout Local Search (BLS) is selected and customized for obtaining high-quality solutions to ASP. A number of ASP problems are solved by the presented BLS and other algorithms in the ASP literature, including simulated annealing, genetic algorithm, memetic algorithm, harmony search, hybrid immune systems-particle swarm optimization, as well as by other variants of local search like iterated local search and multi-start local search. Experimental results and their in-depth statistical analyses show that the BLS outperforms other algorithms by producing the best-known or optimal solutions most of the time.

A modified binary artificial bee colony algorithm for ramp rate constrained unit commitment problem

This paper proposes a new approach based on modified binary artificial bee colony (MBABC) algorithm and dynamic economic dispatch (DED) method for unit commitment problem (UCP). MBABC algorithm is used for committing/decommitting the thermal units, while optimum dispatch solution is determined using DED method. Proposed MBABC algorithm utilizes a new mechanism based on the measure of dissimilarity between binary strings for generating the new binary solutions for UCP. Moreover, in MBABC algorithm, an intelligent scout bee phase is proposed that replaces the abandoned solution with the global best solution. The solution quality achieved by MBABC is enhanced by hybridizing the genetic crossover (GC) that provides the diversified search space. Performance of the proposed MBABC-GC algorithm is tested up to 300 thermal units over 24-hour time interval. The comparison of obtained results with other methods in the literature has confirmed the superiority of the MBABC-GC algorithm in terms of production cost and robustness. Copyright © 2015 John Wiley & Sons, Ltd.

The two-stage hybrid flow shop problem with dedicated machines under release dates and delivery times

The hybrid flow shop scheduling problem has been extensively examined where the main objective has been to improve production efficiency. However, for contemporary manufacturing firms, the due date related performance has gained a significant managerial importance in real life production. Therefore, in this study we consider the integration of both release dates and delivery times for the jobs. We develop several lower bounds and five heuristics. We also implement and test different variants of tabu search. An extensive computational study shows that the proposed methods yield good results within moderate CPU time.

Finding resonant frequencies of microwave cavities through a modified harmony search algorithm

This article describes the computation of the resonant frequencies of a cylindrical applicator through circuital analysis theory, using the unified particle swarm optimisation (UPSO) algorithm and a new variant of the harmony search (HS) algorithm, called self-adjustable bandwidth harmony search (SBHS) algorithm. It was found that SBHS yields more accurate results, but took somewhat longer. Even so, both approaches provided similar results when considering the penalty factor. We also found that as more solutions appear in the frequency range, it seems easier for SBHS to identify them, thus requiring less iterations and computation time. Results showed that this type of models can be solved efficiently using SBHS.

Quality of Service Routing in Manet Using a Hybrid Intelligent Algorithm Inspired by Cuckoo Search.

A hybrid computational intelligent algorithm is proposed by integrating the salient features of two different heuristic techniques to solve a multiconstrained Quality of Service Routing (QoSR) problem in Mobile Ad Hoc Networks (MANETs) is presented. The QoSR is always a tricky problem to determine an optimum route that satisfies variety of necessary constraints in a MANET. The problem is also declared as NP-hard due to the nature of constant topology variation of the MANETs. Thus a solution technique that embarks upon the challenges of the QoSR problem is needed to be underpinned. This paper proposes a hybrid algorithm by modifying the Cuckoo Search Algorithm (CSA) with the new position updating mechanism. This updating mechanism is derived from the differential evolution (DE) algorithm, where the candidates learn from diversified search regions. Thus the CSA will act as the main search procedure guided by the updating mechanism derived from DE, called tuned CSA (TCSA). Numerical simulations on MANETs are performed to demonstrate the effectiveness of the proposed TCSA method by determining an optimum route that satisfies various Quality of Service (QoS) constraints. The results are compared with some of the existing techniques in the literature; therefore the superiority of the proposed method is established.

Adaptive acceleration coefficients for a new search diversification strategy in particle swarm optimization algorithms

The paper presents a novel paradigm of the original particle swarm concept, based on the idea of having two types of agents in the swarm; the “explorers” and the “settlers”, that could dynamically exchange their role in the search process. The explorers’ task is to continuously explore the search domain, while the settlers set out to refine the search in a promising region currently found by the swarm. To obtain this particle task differentiation, the numerical coefficients of the cognitive and social component of the stochastic acceleration as well as the inertia weight were related to the distance of each particle from the best position found so far by the swarm, each of them with a proper distribution over the swarm. This particle task differentiation enhances the local search ability of the particles closer to gbest and improves the exploration ability of the particles as the distance from gbest increases.The originality of this approach is based on the particle task differentiation and on the dynamical adjustment of the particle velocities at each time step on the basis of the current distance of each particle from the best position discovered so far by the swarm.To ascertain the effectiveness of the proposed variant of the PSO algorithm, several benchmark test functions, both unimodal and multi-modal, have been considered and, thanks to its task differentiation concept and adaptive behavior feature, the algorithm has demonstrated a surprising effectiveness and accuracy in identifying the optimal solution.

Special issue on “Variable Neighborhood Search Variants – Recent Applications”

Special issue on “Variable Neighborhood Search Variants – Recent Applications”

Conflict-based search for optimal multi-agent pathfinding

In the multi-agent pathfinding problem (MAPF) we are given a set of agents each with respective start and goal positions. The task is to find paths for all agents while avoiding collisions. Most previous work on solving this problem optimally has treated the individual agents as a single ‘joint agent’ and then applied single-agent search variants of the A* algorithm.In this paper we present the Conflict Based Search (CBS) a new optimal multi-agent pathfinding algorithm. CBS is a two-level algorithm that does not convert the problem into the single ‘joint agent’ model. At the high level, a search is performed on a Conflict Tree (CT) which is a tree based on conflicts between individual agents. Each node in the CT represents a set of constraints on the motion of the agents. At the low level, fast single-agent searches are performed to satisfy the constraints imposed by the high level CT node. In many cases this two-level formulation enables CBS to examine fewer states than A* while still maintaining optimality. We analyze CBS and show its benefits and drawbacks.Additionally we present the Meta-Agent CBS (MA-CBS) algorithm. MA-CBS is a generalization of CBS. Unlike basic CBS, MA-CBS is not restricted to single-agent searches at the low level. Instead, MA-CBS allows agents to be merged into small groups of joint agents. This mitigates some of the drawbacks of basic CBS and further improves performance. In fact, MA-CBS is a framework that can be built on top of any optimal and complete MAPF solver in order to enhance its performance. Experimental results on various problems show a speedup of up to an order of magnitude over previous approaches.

Fuzzy adaptive imperialist competitive algorithm for global optimization

Most of the parameters in meta-heuristic algorithms need precise tuning before applying into the optimization problems. The quality of solutions obtained by meta-heuristic algorithms is highly dependent on how efficiently its parameters are tuned by the users. Setting the parameters is not straightforward, and there is no definitive way to determine their appropriate values other than trial and error, which is an inefficient and time-consuming process. Deviation angle parameter in the imperialist competitive algorithm (ICA) plays an important role as driver of search diversification and intensification. In this paper, we propose a fuzzy adaptive imperialist competitive algorithm (FAICA), where the deviation parameter is adaptively adjusted using a fuzzy controller to optimally control diversification and intensification through the search process. A set of benchmark function tests has been carried out to demonstrate the effectiveness and robustness of the proposed algorithm. The results of experiments are compared with the original ICA and other five well-known algorithms. The experimental results exhibit the merit of achieving balance between intensification and diversification in the search which indicates superiority of FAICA over ICA.