Search

In recent decades, researches on optimizing the parameter of the artificial neural network (ANN) model has attracted significant attention from researchers. Hybridization of superior algorithms helps improving optimization performance and capable of solving complex applications. As a traditional gradient-based learning algorithm, ANN suffers from a slow learning rate and is easily trapped in local minima when training techniques such as gradient descent (GD) and back-propagation (BP) algorithm are used. The characteristics of randomization and selection of the best or near-optimal solution of metaheuristic algorithm provide an effective and robust solution; therefore, it has always been used in training of ANN to improve and overcome the above problems. New metaheuristic algorithms are proposed every year. Therefore, the review of its latest developments is essential. This article attempts to summarize the metaheuristic algorithms which have been proposed from the year 1975 to 2020 from various journals, conferences, technical papers, and books. The comparison of the popularity of the metaheuristic algorithm is presented in two time frames, such as algorithms proposed in the recent 20 years and those proposed earlier. Then, some of the popular metaheuristic algorithms and their working principle are reviewed. This article further categorizes the latest metaheuristic search algorithm in the literature to indicate their efficiency in training ANN for various industry applications. More and more researchers tend to develop new hybrid optimization tools by combining two or more metaheuristic algorithms to optimize the training parameters of ANN. Generally, the algorithm's optimal performance must be able to achieve a fine balance of their exploration and exploitation characteristics. Hence, this article tries to compare and summarize the properties of various metaheuristic algorithms in terms of their convergence rate and the ability to avoid the local minima. This information is useful for researchers working on algorithm hybridization by providing a good understanding of the convergence rate and the ability to find a global optimum. [ABSTRACT FROM AUTHOR]

In the presented paper, the functioning and the results of the work of two metaheuristic algorithms, namely cuckoo search algorithm (CS) and firefly algorithm (FA), are described using the apparatus of generalized nets (GNs), which is an appropriate and efficient tool for describing the essence of various optimization methods. The two developed GN-models mimic the optimization processes based on the nature of cuckoos and fireflies, respectively. The proposed GN-models execute the two considered metaheuristic algorithms conducting basic steps and performing optimal search. Building upon these two GN-models, a universal GN-model is constructed that can be used for describing and simulating both the CS and the FA by setting different characteristic functions of the GN-tokens. Moreover, the universal GN-model itself can be transformed to each of the herewith presented GN-models by applying appropriate hierarchical operators. In order to validate the proposed universal GN-model, numerical experiments are performed for the operating of the universal GN-model (CS and FA) on benchmark mathematical functions. The obtained results are compared with the results of the GN-model of CS, GN-model of FA, as well as the results of the standard CS and FA. [ABSTRACT FROM AUTHOR]

Nowadays, wireless sensor networks (WSNs) are considered in many fields of application. In this paper, we study how to efficiently deploy relay nodes into previously established static WSNs, with the purpose of optimising two relevant factors for the industry: average energy consumption of the sensors and average sensitivity area provided by the network. This is the so-called relay node placement problem, which is a known NP-hard optimisation problem in the literature. With the purpose of tackling this multiobjective (MO) optimisation problem, we consider two different approaches of the trajectory algorithm MO-VNS, assuming a wide range of stop conditions. Two additional standard genetic algorithms are included in this study, NSGA-II and SPEA2, which belong to evolutionary algorithms. The aim is to analyse the behaviour of MO-VNS compared to traditional methodologies. To this end, the four metaheuristics are applied to solve a freely available data set. The results obtained are analysed following a widely accepted statistical methodology and considering three MO quality metrics: hypervolume, set coverage, and attainment surface. After studying the results, we conclude that MO-VNS provides better performance than the standard algorithms NSGA-II and SPEA2. Moreover, we verify that the addition of relay nodes is a good way to optimise traditional WSNs. [ABSTRACT FROM AUTHOR]