meta-heuristic algorithms

Optimization plays a crucial role in enhancing productivity within the industry. Employing this technique can lead to a reduction in system costs. There exist various efficient methods for optimization, each with its own set of advantages and disadvantages. Meanwhile, meta-heuristic algorithms offer a viable solution for achieving the optimal working point. These algorithms draw inspiration from nature, physical relationships, and other sources. The distinguishing factors between these methods lie in the accuracy of the final optimal solution and the speed of algorithm execution. The superior algorithm provides both precise and rapid optimal solutions. This paper introduces a novel agricultural-inspired algorithm named Elymus Repens Optimization (ERO). This optimization algorithm operates based on the behavioral patterns of Elymus Repens under cultivation conditions. Elymus repens is inclined to move to areas with more suitable conditions. In ERO, exploration and exploitation are carried out through Rhizome Optimization Operator and Stolon Optimization Operators. These two supplementary activities are used to explore the problem space. The potent combination of these operators, as presented in this paper, resolves the challenges encountered in previous research related to speed and accuracy in optimization issues. After the introduction and simulation of ERO, it is compared with popular search algorithms such as Gravitational Search Algorithm (GSA), Grey Wolf Optimizer (GWO), Particle Swarm Optimization (PSO), and Firefly Algorithm (FA). The solution of 23 benchmark functions demonstrates that the proposed algorithm is highly efficient in terms of accuracy and speed.

Medical image registration plays an important role in many clinical applications, including the detection and diagnosis of diseases, planning of therapy, guidance of interventions. Multimodal medical image registration is the process of overlapping two or more images taken from the same scene by different modalities and different sensors. Intensity-based methods are widely used in multimodal medical image registration, these techniques register different modality images that have the same content by optimal transformation. The estimation of the optimal transformation requires the optimization of a similarity metric between the images. Recently, various optimization algorithms have been presented that the selection of appropriate optimization algorithms is very important in determining the optimal transformation parameter. The Social Spider Optimization (SSO) algorithm is one of the meta-heuristic methods that prevents premature convergence. In this paper, medical image registration technique is suggested based on the SSO algorithm. The Mutual Information (MI), Normalization of Mutual Information (NMI), and Sum of Squared Differences (SSD) are used separately as cost function (objective function) and the performance of each of these functions is checked in multimodal medical image registration. The simulation results on Brain Web data set affirm the suggested method outperforms classical registration methods in terms of convergence rate, execution time.

Optimization problems are becoming more complicated, and their resource requirements are rising. Real-life optimization problems are often NP-hard and time or memory consuming. Nature has always been an excellent pattern for humans to pull out the best mechanisms and the best engineering to solve their problems. The concept of optimization seen in several natural processes, such as species evolution, swarm intelligence, social group behavior, the immune system, mating strategies, reproduction and foraging, and animals’ cooperative hunting behavior. This paper proposes a new Meta-Heuristic algorithm for solving NP-hard nonlinear optimization problems inspired by the intelligence, socially, and collaborative behavior of the Qashqai nomad’s migration who have adjusted for many years. In the design of this algorithm uses population-based features, experts’ opinions, and more to improve its performance in achieving the optimal global solution. The performance of this algorithm tested using the well-known optimization test functions and factory facility layout problems. It found that in many cases, the performance of the proposed algorithm was better than other known meta-heuristic algorithms in terms of convergence speed and quality of solutions. The name of this algorithm chooses in honor of the Qashqai nomads, the famous tribes of southwest Iran, the Qashqai algorithm.

Feature selection is the process of identifying relevant features and removing irrelevant and repetitive features with the aim of observing a subset of features that describe the problem well and with minimal loss of efficiency. One of the feature selection approaches is using optimization algorithms. This work provides a summary of some meta-heuristic feature selection methods proposed from 2018 to 2021 that were designed and implemented on a wide range of different data. The results of the study showed that some meta-heuristic algorithms alone cannot perfectly solve the feature selection problem on all types of datasets with an acceptable speed. In other words, depending on dataset, a special meta-heuristic algorithm should be used.

In this paper, the performance of meta-heuristic algorithms is compared using statistical analysis based on new criteria (powerfulness and effectiveness). Due to the large number of meta-heuristic methods reported so far, choosing one of them by researchers has always been challenging. In fact, the user does not know which of these methods are able to solve his complex problem. In this paper, in order to compare the performance of several methods from different categories of meta-heuristic methods new criteria are proposed. In fact, by using these criteria, the user is able to choose an effective method for his problem. For this reason, statistical analysis is conducted on each of these methods to clarify the application of each of these methods for the users. Also, powerfulness and effectiveness criteria are defined to compare the performance of the meta-heuristic methods to introduce suitable substrate and suitable quantitative parameters for this purpose. The results of these criteria clearly show the ability of each method for different applications and problems.