This paper presents a concise literature review of metaheuristic algorithms in feature selection, spanning publications from 2019 to 2023. The study acknowledges the role of wrapper methods and metaheuristics, noting their ability to yield enhanced results. It highlights the prevalent use of Particle Swarm Optimization, Grey Wolf Optimizer, and Genetic Algorithm in this context. Additionally, the research explores trends and approaches in binarization within metaheuristics, distinguishing between straightforward binarization and more elaborate methods. A central theme of the study is the investigation of hybridization in metaheuristics, demonstrating the use and integration of multiple algorithms in search of improvement in performance. The paper also discusses strategies for refining metaheuristic performance, such as chaotic maps and local search. It examines the emerging domain of multi-objective metaheuristics, which is particularly relevant for addressing real-world problems with competing objectives. The study highlights the dynamic and innovative potential of metaheuristic-based feature selection methodologies.

Technological advances and the digitization of information have allowed us to obtain a large amount of data from different processes such as medicine, commerce, mining, among others. All this data has been used by different researchers in machine learning techniques to accelerate the decision making process of professionals. Machine learning techniques are very sensitive to data, so it is necessary to perform a cleaning to remove irrelevant and redundant information. This information removal is known as the feature selection problem. This paper presents the Pendulum Search Algorithm applied to solve the feature selection problem. Since the Pendulum Search Algorithm is a metaheuristic designed for continuous optimization problems, a binarization process is performed using the two-step technique. Preliminary results indicate that our proposal obtains competitive results compared to other metaheuristics extracted from the literature that solve well-known benchmarks.

Constraint Programming (CP) is a well-established area in AI as a programming paradigm for modelling and solving discrete optimization problems, and it has been been successfully applied to tackle the on-line job dispatching problem in HPC systems including those running modern applications. The limitations of the available CP-based job dispatchers may hinder their practical use in today's systems that are becoming larger in size and more demanding in resource allocation. In an attempt to bring basic AI research closer to a deployed application, we present a new CP-based on-line job dispatcher for modern HPC systems and applications. Unlike its predecessors, our new dispatcher tackles the entire problem in CP and its model size is independent of the system size. Experimental results based on a simulation study show that with our approach dispatching performance increases significantly in a large system and in a system where allocation is nontrivial.

Today, natural resources are more scarce than ever, so we must make good use of them. To achieve this goal, we can use metaheuristic optimization tools as an alternative to achieve good results in a reasonable amount of time. The present work focuses on the use of adaptive techniques to facilitate the use of this type of tool to obtain good functional parameters. We use a constructive metaheuristic algorithm called Intelligent Water Drops to solve the set covering problem. To demonstrate the efficiency of the proposed method, the obtained results were compared with the standard version using the same initial configuration for both algorithms. Additionally, the Kolmogorov-Smirnov-Lilliefors, Wilcoxon signed-rank and Violin chart tests were applied to statistically validate the results, which showed that metaheuristics with autonomous search have a better behavior than do standard algorithms.

The cell formation problem is a classic optimization problem devoted to the manufacturing industry. Such a problem proposes to divide a manufacturing plant in a set of cells, where each cell is composed of machines which in turn process product parts. The goal is to design a plant division in such a way the need for part interchange among cells is minimized. The idea is to reduce cost and increase productivity. In this paper, we propose different variations of the original migrating birds optimization algorithm for solving this problem. In particular, we propose two different leader exchange procedures and three different neighboring solution generations. We illustrate interesting results by solving well-known instances considering the group efficiency as optimization criterion in contrast to previous work done on this metaheuristic.

The Set Covering Problem (SCP) is one of the classical Karp's 21 NP-complete problems. Although this is a traditional optimization problem, we find many papers assuming metaheuristics for solving the SCP in the current literature. However, while the SCP is a discrete problem, most metaheuristics are defined for solving continuous optimization problems, specially Swarm Intelligence Algorithms (SIAs). Hence, such algorithms should be adapted for working on the discrete scope, but most authors did not perform any study to select a concrete binarization approach. This situation might lead to the conclusion that selecting a concrete binarization technique does not influence the behavior of the algorithm, but rather the general approach of the metaheuristic. This circumstance led us to write this paper focusing on the inherent difficulty in binarization of metaheuristics designed for continuous optimization, when solving a discrete optimization problem, concretely the SCP. To this end, we consider a recent SIA inspired by the behavior of cats and adapted to the discrete scope, which is called Binary Cat Swarm Optimization (BCSO). We replace the binarization technique assumed in the original BCSO by forty different approaches from the current literature. The results obtained while solving a standard SCP benchmark are analyzed through a widely accepted statistical method, concluding that it is crucial to select an adequate binarization approach to ensure that the solving algorithm reaches its full potential. Thus, the task of adapting a metaheuristic to the discrete scope is not a simple matter and should be carefully studied. To this end and as a result of this study, we give some recommendations to perform this task.

In operation research and optimization area, Autonomous Search is a technique that provides the solver the auto-adaptive capability, during search process. This technique aims to improve performance in the exploration of search tree, updating the enumeration strategy online. This task is controlled by a choice function (CF) which decides, based on performance indicators given from the solver, how the strategy must be updated. The relevance of indicators is handled via back hole algorithm, inspired on natural phenomenon that occurs in outer space. If choice function exhibits a poor performance, the strategy is replacement and solver continue exploring the search tree under new enumeration strategy. In this paper, we present an evaluation of the impact and efficient using 16 different carefully constructed choice functions. We employ as test bed a set of well-known constrain satisfaction problems. Encouraging experimental results are obtained in order to show which using choice functions is highly efficient, if want to control the search process, online way.

BACKGROUND: Metaheuristics are widely used to solve large combinatorial optimization problems in bioinformatics because of the huge set of possible solutions. Two representative problems are gene selection for cancer classification and biclustering of gene expression data. In most cases, these metaheuristics, as well as other non-linear techniques, apply a fitness function to each possible solution with a size-limited population, and that step involves higher latencies than other parts of the algorithms, which is the reason why the execution time of the applications will mainly depend on the execution time of the fitness function. In addition, it is usual to find floating-point arithmetic formulations for the fitness functions. This way, a careful parallelization of these functions using the reconfigurable hardware technology will accelerate the computation, specially if they are applied in parallel to several solutions of the population. RESULTS: A fine-grained parallelization of two floating-point fitness functions of different complexities and features involved in biclustering of gene expression data and gene selection for cancer classification allowed for obtaining higher speedups and power-reduced computation with regard to usual microprocessors. CONCLUSIONS: The results show better performances using reconfigurable hardware technology instead of usual microprocessors, in computing time and power consumption terms, not only because of the parallelization of the arithmetic operations, but also thanks to the concurrent fitness evaluation for several individuals of the population in the metaheuristic. This is a good basis for building accelerated and low-energy solutions for intensive computing scenarios.

In Castro et al. [Castro, C., Monfroy, E., Figueroa, C., and Meneses, R. (2005), ‘An Approach for Dynamic Split Strategies in Constraint Solving’, in Proceedings of MICAI 2005 (Vol. 3789), LNAI, Berlin: Springer, pp. 162–174] a framework for adaptive enumeration strategies and meta-backtracks for a propagation-based constraint solver has been studied. Here, we extend this framework in order to trigger some functions of a solver, or of a hybrid solver to respond to some observations of the solving process. We can also simply design adaptive hybridisation strategies by just changing some rules of the update component of our framework. We experiment with this framework on a hybrid Branch and Bound + propagation solver in which propagation can be triggered w.r.t. some observations of the solving process. The results show that some phases of propagation are not only beneficial to the Branch and Bound algorithm, but also that propagation is too costly to be executed at each node of the search tree. The hybridisation strategies are thus crucial in order to decide when to perform the or not propagation.

Constraint programming can definitely be seen as a model-driven paradigm. The users write programs for modeling problems. These programs are mapped to executable models to calculate the solutions. This paper focuses on efficient model management (definition and transformation). From this point of view, we propose to revisit the design of constraint-programming systems. A model-driven architecture is introduced to map solving-independent constraint models to solving-dependent decision models. Several important questions are examined, such as the need for a visual highlevel modeling language, and the quality of metamodeling techniques to implement the transformations. A main result is the s-COMMA platform that efficiently implements the chain from modeling to solving constraint problems

This paper presents the first implementation of COMMA, a new solver independent language for modeling constraint- based problems. The combination of a constraint language with an object-oriented framework represents the base of the core of COMMA. Extension capabilities have also been included with the aim of tackling a wide range of applications from combinatorial to continuous problems. A COMMA compiler has been implemented through a three layered architecture including a dynamic parsing system for handling efficiently the mapping process. In particular, COMMA models can be translated to different solvers, currently to ECLiPSe and Gecode/J.