Search

The lemur optimizer (LO) is a new metaheuristic algorithm that is inspired by the two primary locomotive patterns of the lemurs. However, LO suffers from improper balance between exploration and exploitation phases and sub-optimal solution stagnation due to poor exploration capability. To address these limitations, an improved lemur optimizer (ILO) with dynamic scaling factors and Gaussian mutation is introduced in this paper. The dynamic scaling factors balance the exploration and exploitation transition by dynamically adjusting their values throughout the optimization process. The proposed algorithm incorporates Gaussian mutation to induce diversity in the population and hence improve the exploration capability. The efficiency of the proposed approach is tested on 33 functions including classical benchmark and CEC-2019 test functions, and a comprehensive comparison is made with six other well-known metaheuristic algorithms. The simulation and statistical results indicate superior performance of ILO in achieving optimal solutions compared to the competing algorithms. The algorithm is further employed to address two constrained classical engineering design problems, demonstrating its robustness. [ABSTRACT FROM AUTHOR]

Since the past few years, several metaheuristic algorithms, inspired by the natural processes, have been introduced to solve different complex optimization problems. Studying and comparing the convergence, computational burden and statistical significance of those metaheuristics are helpful for future algorithmic development and their applications. This paper focuses on comparing the optimization performance of classical dragonfly algorithm (DA) and its seven different variants, i.e., hybrid memory-based dragonfly algorithm with differential evolution (DADE), quantum-behaved and Gaussian mutational dragonfly algorithm (QGDA), memory-based hybrid dragonfly algorithm (MHDA), chaotic dragonfly algorithm (CDA), biogeography-based Mexican hat wavelet dragonfly algorithm (BMDA), hybrid Nelder–Mead algorithm and dragonfly algorithm (INMDA) and hybridization of dragonfly algorithm and artificial bee colony (HDA) while solving 80 CEC-2021 benchmark problems. It is observed that the convergence rates of different variants of DA algorithm vary, and the corresponding computational times for such variations are also evaluated. This paper finally ranks DA and its variants according to their convergence efficiency and Friedman test. The DADE, QGDA, BMDA and DA evolve out as the most efficient algorithms for solving the considered CEC-2021 benchmark problems. [ABSTRACT FROM AUTHOR]

The marine predator algorithm is a new nature-inspired metaheuristic algorithm that mimics biological interaction between marine predators and prey. It has been also stated from the literature that this algorithm can solve many real-world optimization problems which made it a new popular optimization technique for the researchers. However, there is still a deficiency in the marine predator algorithm such as the inability to produce a diverse initial population with high productivity, lack of quick escaping of the local optimization, and lack of widely and broadly exploration of the search space. In the present study, a developed version of this algorithm is proposed based on the opposition-based learning method, chaos map, self-adaptive of population, and switching between exploration and exploitation phases. The simulations are performed using MATLAB environment on standard test functions including CEC-06 2019 tests and a real-world optimization problem based on PID control applied to a DC motor to evaluate the performance of the suggested algorithm. The simulation results are compared with the original marine predator algorithm and five state-of-the-art optimization algorithms namely Particle Swarm Optimization, Grasshopper Optimization Algorithm, JAYA Algorithm, Equilibrium optimizer Algorithm, Whale Optimization Algorithm, Differential Search Algorithm, and League Championship Algorithm. Eventually, the simulation results proved that the suggested algorithm has better results compared with other algorithms for the studied case studies. [ABSTRACT FROM AUTHOR]

Feature transformation methods can be used effectively to improve the performance of the machine learning techniques, such as classification and clustering, via enhancing the discrimination among the samples. Additionally, binary representation of features may be necessary in several circumstances. For example, distributed systems may have limited bandwidth, storage, and energy that necessitate rough quantization of the measurements. This paper proposes a random projection-based feature transformation method that maps the data points from the original space into a new binary space. In the proposed transformation method, the random projection process is formulated as an optimization problem. A modified chaotic version of the sine–cosine algorithm is used to find the optimal random projector. The performance of the proposed transformation method is validated using 15 standard UCI benchmark datasets against the linear discernment analysis. Moreover, it is compared to other metaheuristic-based feature reduction methods including the gray wolf optimizer, the genetic algorithm, and the antlion optimizer. The obtained results assure the superiority of the proposed transformation method in terms of classification accuracy and storage requirements. [ABSTRACT FROM AUTHOR]

In this paper, a new metaheuristic algorithm called JAYA algorithm has been adapted for feature selection. Feature selection is a typical problem in machine learning and data mining domain concerned with determining the subset of high discriminative features from the irrelevant, noisy, redundant, and high-dimensional features. JAYA algorithm is initially proposed for continuous optimization. Due to the binary nature of the feature selection problem, the JAYA algorithm is adjusted using sinusoidal (i.e., S-shape) transfer function. Furthermore, the mutation operator controlled by adaptive mutation rate ( R m ) parameter is also utilized to control the diversity during the search. The proposed binary JAYA algorithm with adaptive mutation is called BJAM algorithm. The performance of BJAM algorithm is tested using 22 real-world benchmark datasets, which vary in terms of the number of features and the number of instances. Four measures are used for performance analysis: classification accuracy, number of features, fitness values, and computational times. Initially, a comparison between binary JAYA (BJA) algorithm and the proposed BJAM algorithm is conducted to show the effect of the mutation operator in the convergence behavior. After that, the results produced by the BJAM algorithm are compared against those yielded by ten state-of-the-art methods. Surprisingly, the proposed BJAM algorithm is able to excel other comparative methods in 7 out of 22 datasets in terms of classification accuracy. This can lead to the conclusion that the proposed BJAM algorithm is an efficient algorithm for the problems belonging to the feature selection domain and is pregnant with fruitful results. [ABSTRACT FROM AUTHOR]

Due to the lack of experience and prior knowledge, the selection of the most informative features has become one of the challenging problems in many applications. Recently, many metaheuristic algorithms have widely used to solve the feature selection problem for classification tasks. In this paper, the chaotic atom search optimization (CASO) that integrates the chaotic maps into atom search optimization (ASO) is applied for wrapper feature selection. Twelve different chaotic maps are used to adjust the parameter of CASO through the optimization process, which is beneficial for enhancing the convergence rate and improving the efficiency of ASO algorithm. In this study, twenty benchmark datasets acquired from the UCI machine learning repository are used to validate the performance of CASO in feature selection. Several state-of-the-art metaheuristic algorithms are adopted to examine the efficacy and effectiveness of the proposed approach. Our results indicated that the Logistic-Tent map was the most suitable chaotic map to boost the performance of CASO. The experimental result shows the capability of CASO not only in finding the optimal solution but also in significantly improving the prediction accuracy and reducing the number of features. [ABSTRACT FROM AUTHOR]

Nature-inspired metaheuristic algorithms have significantly influenced the optimization problems, and there has been a great success in finding optimum solutions. The main aim of this paper is to introduce one such technique, runner-root algorithm (RRA), to the antenna community. RRA is based on the concept of plant intelligence to solve an optimization problem resembling a scenario where a plant finds the best location for its survival. The main advantage of RRA over other algorithms is the execution of exploration (global search) and exploitation (local search) phases independently with a built-in mechanism of reinitialization to overcome trapping in local minima. Hence, attempt has been made for the first time to apply this algorithm to optimize excitation current amplitudes and positions of antenna elements for linear array configurations with the objective of suppressing the sidelobe levels and/or placing nulls at prescribed locations referring to the radiation patterns. Different case studies have illustrated that RRA is capable of performing linear antenna array optimization with promising results in comparison with other well-established metaheuristic algorithms. And hence such demonstration has identified RRA as a potential candidate in the optimization domain. [ABSTRACT FROM AUTHOR]

The optimization of continuous space poses a great challenge among the scientific community. When the objective function is nonlinear, the choices of direct search spaces are preferred over the other methods. The use of the hybrid algorithm for these types of optimization is becoming increasingly popular. This study introduced a self-adaptation procedure in a single objective hybrid algorithm and its application for unconstrained and constrained optimization test functions. This single objective hybrid algorithm is based on two popular metaheuristic algorithms, namely the cuckoo search and covariance matrix adaptation evolution strategy. Self-adaptation is a popular way of parameter selection and has a significant place in the computing field. The adaptation is introduced in two significant parameters of this algorithm. Five metaheuristic algorithms, namely cuckoo search, covariance matrix adaptation evolution strategy, particle swarm intelligence, firefly algorithm, and the newly introduced self-adapted single objective hybrid algorithm, were analyzed using unconstrained (unimodal and multimodal) and constrained benchmark test functions. An encouraging performance of this proposed algorithm for unconstrained and constrained test functions was observed. [ABSTRACT FROM AUTHOR]