Article
Computer Science, Artificial Intelligence
Lei Zhao, Rui Li, Jianda Han, Jianlei Zhang
Summary: This article proposes a framework for multidifferent-target search in unknown environments based on swarm intelligence. The framework introduces the idea of distributed model predictive control in the target search method and enhances the robot's path prediction ability using a hierarchical prediction strategy. Compared to existing methods, this strategy significantly improves the functionality and success rate of multidifferent-target search in unknown complex obstacle environments. The article also presents two effective efforts to reduce computational complexity and speed up decision making.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Mathematical & Computational Biology
Sunil Kumar Prabhakar, Young-Gi Ju, Harikumar Rajaguru, Dong-Ok Won
Summary: This study proposes a novel method combining sparse representation and Swarm Intelligence techniques for EEG signal classification. Experimental results show that the method achieves a maximum classification accuracy of 98.94% when combined with deep learning and 95.70% when combined with SI-based HMM method.
FRONTIERS IN COMPUTATIONAL NEUROSCIENCE
(2022)
Article
Computer Science, Information Systems
Abhishek Sharma, Abhinav Sharma, Moshe Averbukh, Vibhu Jately, Brian Azzopardi
Summary: The article proposed a new hybrid algorithm to optimize photovoltaic cell parameters, achieving good performance and obtaining minimum root mean square error under low irradiation conditions.
Article
Chemistry, Physical
Sida Zhou, Xinhua Liu, Yang Hua, Xinan Zhou, Shichun Yang
Summary: This article introduces a coupled hybrid adaptive particle swarm optimization-hybrid simulated annealing algorithm for precise parameter identification, which is validated on three different types of batteries and shows excellent consistency between simulation results and experimental data.
JOURNAL OF POWER SOURCES
(2021)
Article
Mathematics
Khizer Mehmood, Naveed Ishtiaq Chaudhary, Zeshan Aslam Khan, Muhammad Asif Zahoor Raja, Khalid Mehmood Cheema, Ahmad H. Milyani
Summary: This study investigates the parameter estimation of the CAR model using a novel AO algorithm, demonstrating the accuracy, convergence, and robustness of the AO under various noise levels.
Article
Computer Science, Interdisciplinary Applications
Chen Wang, Wenxi Kuang, Minqiang Gu, Zhun Fan
Summary: This study proposes a distributed algorithm that enables agents' adaptive grouping and entrapment of multiple targets via automatic decision making, smooth flocking, and well-distributed entrapping. The study introduces an agent distributed decision framework where agents make their own decisions on which targets to surround based on environmental information. Additionally, a modified Vicsek model is proposed to allow agents to smoothly change formations, adapt to the environment, and effectively entrap the target. The performance of the proposed method is validated through simulation and physical experiments.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2023)
Review
Automation & Control Systems
Shanling Dong, Meiqin Liu, Zheng-Guang Wu
Summary: In recent years, there has been a significant amount of research on the problems of asynchronous control and filtering for Markov jump systems (MJSs). The use of hidden Markov model (HMM) allows for the modeling of the asynchronous situation between the original MJSs and the controller/filter. This survey reviews the recent development of HMM-based asynchronous controller and filter design for different types of MJSs, such as linear MJSs, fuzzy MJSs, semi-MJSs, and 2D MJSs. The conclusion summarizes the findings and discusses potential future research directions for MJSs.
INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE
(2023)
Article
Physics, Multidisciplinary
Ozgur Danisman, Umay Uzunoglu Kocer
Summary: Hidden Markov models are commonly used for modeling probabilistic structures with latent variables. They assume that observation symbols are conditionally independent and identically distributed, but in practice, this assumption may not always hold. The proposed model introduces a first-order Markov dependency between the current pair of hidden state-emitted observation symbol and the previous pair, which can better capture possible dependencies in real-life scenarios.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2021)
Review
Computer Science, Artificial Intelligence
Marcelo Gomes Pereira de Lacerda, Luis Filipe de Araujo Pessoa, Fernando Buarque de Lima Neto, Teresa Bernarda Ludermir, Herbert Kuchen
Summary: This paper presents a systematic literature review on general parameter control for evolutionary and swarm-based algorithms, with a total of 4449 studies retrieved and only 15 fully analyzed and discussed. It is the first literature review in this field and one of the very few systematic reviews on parameter adjustment for those algorithms.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Information Systems
Debina Laishram, Themrichon Tuithung
Summary: This study presents J-HMMSteg, an adaptive and secure JPEG image steganography technique that embeds data with minimal distortion. J-HMMSteg utilizes block-wise analysis, statistical feature construction, Hidden Markov Model (HMM) creation, and a maximum likelihood embedder to achieve data embedding. Experimental results demonstrate that J-HMMSteg performs well in terms of imperceptibility, robustness against RS steganalysis, and enhanced security.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Marcelo Gomes Pereira de Lacerda, Hugo de Andrade Amorim Neto, Teresa Bernarda Ludermir, Herbert Kuchen, Fernando Buarque de Lima Neto
Summary: This article presents a parameterless out-of-the-box population size control method for evolutionary and swarm-based algorithms in single objective bound constrained real-parameter numerical optimization. It incrementally changes the velocity of population change based on the stagnation of fitness and utilizes a mechanism inspired by evolutionary algorithms for individual removal/addition to effectively change the population size. Experimental results demonstrate that the controller is compatible and performs well in various scenarios.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Information Systems
Junji Jiang, Likang Wu, Hongke Zhao, Hengshu Zhu, Wei Zhang
Summary: Stock movement forecasting is often treated as a sequence prediction task using time series data. While deep learning models have been increasingly employed for fitting dynamic stock time series, few of them have focused on understanding the internal dynamics of the market system. To address this, the proposed HMM-ALSTM framework integrates the Hidden Markov Model (HMM) into the deep learning process, allowing for the discovery of hidden states and patterns that contribute to the stock time series data.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Physics, Multidisciplinary
Wei Zhou, Limin Wang, Xuming Han, Yizhang Wang, Yufei Zhang, Zhiyao Jia
Summary: The DBSCAN algorithm is sensitive to the neighborhood radius and noise points, making it difficult to quickly obtain accurate results. To address this, we propose an adaptive DBSCAN method using the chameleon swarm algorithm. By iteratively optimizing the evaluation index of DBSCAN, we can find the best neighborhood radius and clustering result. Additionally, the algorithm solves the problem of over-identification of noise points by utilizing the theory of deviation in nearest neighbor searches.
Article
Computer Science, Artificial Intelligence
Min Xue, Huaicheng Yan, Hao Zhang, Jun Sun, Hak-Keung Lam
Summary: This article addresses the issue of imperfect premise matching and asynchronous behavior in H-infinity output tracking control for Takagi-Sugeno fuzzy Markov jump systems. A hidden Markov model is utilized to capture the asynchronous phenomenon between system and controller modes, with packet loss described by a stochastic variable. Novel Lyapunov function is employed to derive stability criteria and develop an asynchronous control scheme with H-infinity tracking performance, validated through two examples.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2021)
Article
Energy & Fuels
Husheng Wu, Qiang Peng, Meimei Shi, Lining Xing, Shi Cheng
Summary: In this study, a novel algorithm called DCWPA is proposed for rapid and accurate identification of solar cell model parameters. By using chaotic map sequence to initialize the population and adopting walking direction mechanism based on drunk walking model and adaptive walking step size, the algorithm shows higher performance in terms of identification accuracy and algorithm optimization. Experimental results verify the effectiveness and superiority of the improved algorithm in solar cell parameter identification.
Article
Mathematics
Francisco Gonzalez, Ricardo Soto, Broderick Crawford
Summary: In this study, a hybrid approach combining stochastic fractal search algorithm with opposition-based learning was used to design substitution boxes with high nonlinearity. The proposed approach outperformed other methods based on metaheuristics and chaotic schemes in terms of performance evaluation metrics.
Article
Mathematics
Emanuel Vega, Ricardo Soto, Pablo Contreras, Broderick Crawford, Javier Pena, Carlos Castro
Summary: Population-based approaches offer new search strategies for optimization problems. This work proposes a hybrid architecture that intelligently balances population size by using learning components and statistical modeling methods. It demonstrates the viability and effectiveness of the approach through solving benchmark functions and the multidimensional knapsack problem.
Article
Computer Science, Artificial Intelligence
Emanuel Vega, Ricardo Soto, Broderick Crawford, Javier Pena, Pablo Contreras, Carlos Castro
Summary: This paper explores the ability of metaheuristics to autonomously predict population size and termination criteria, and presents interesting results in the experiments.
APPLIED SOFT COMPUTING
(2022)
Article
Operations Research & Management Science
Alejandro Fernandez Gil, Eduardo Lalla-Ruiz, Mariam Gomez Sanchez, Carlos Castro
Summary: The Cumulative Vehicle Routing Problem with Time Windows (CumVRP-TW) is a vehicle routing variant that minimizes a cumulative cost function while respecting customers' time windows constraints. Mathematical models are proposed for soft and hard time windows constraints, allowing violations in the soft case subject to penalization. A matheuristic approach combining features of the Greedy Randomized Adaptive Search Procedure (GRASP) and the optimization model is proposed to solve the problem. Computational results show that the mathematical formulations provide optimal solutions for small scenarios, but the proposed matheuristic algorithm outperforms the exact solver for medium and large scenarios and reduces fuel consumption and carbon emissions when soft time windows are allowed to be violated.
ANNALS OF OPERATIONS RESEARCH
(2023)
Editorial Material
Chemistry, Multidisciplinary
Juan A. A. Gomez-Pulido, Young Park, Ricardo Soto, Jose M. Lanza-Gutierrez
APPLIED SCIENCES-BASEL
(2023)
Review
Mathematics
Marcelo Becerra-Rozas, Jose Lemus-Romani, Felipe Cisternas-Caneo, Broderick Crawford, Ricardo Soto, Gino Astorga, Carlos Castro, Jose Garcia
Summary: This study is a continuation of research on the binarization of continuous metaheuristics for solving binary-domain combinatorial problems. By analyzing 512 publications from 2017 to January 2022, the authors provide a comprehensive overview of the various ways to binarize this type of metaheuristics. The findings offer a theoretical foundation for novice researchers and expert researchers in the field of combinatorial optimization using metaheuristic algorithms, and highlight the impact of binarization mechanism on the performance of metaheuristic algorithms. The study emphasizes that there is no single general technique for efficient binarization, but rather multiple forms with different performances.
Article
Mathematics
Jose Garcia, Andres Leiva-Araos, Broderick Crawford, Ricardo Soto, Hernan Pinto
Summary: This paper examines the impact of solution initialization methods on the performance of a hybrid algorithm applied to the set union knapsack problem (SUKP). The weighted method outperforms random and greedy methods, demonstrating its effectiveness in improving algorithm performance. The results are compared with other metaheuristics that have previously solved SUKP, further showcasing the favorable performance of the weighted method.
Article
Mathematics, Applied
Rodrigo Olivares, Ricardo Soto, Broderick Crawford, Victor Rios, Pablo Olivares, Camilo Ravelo, Sebastian Medina, Diego Nauduan
Summary: This paper presents adaptive parameter control methods utilizing reinforcement learning in the particle swarm algorithm. The integration of Q-learning into the optimization algorithm allows for parameter adjustments during the run, enabling the algorithm to dynamically learn and adapt to the problem and its context. Experimental evaluation using instances of the NP-hard multidimensional knapsack problem demonstrates significant improvement in solution quality compared to the native version of the algorithm.
Article
Engineering, Multidisciplinary
Jose Lemus-Romani, Broderick Crawford, Felipe Cisternas-Caneo, Ricardo Soto, Marcelo Becerra-Rozas
Summary: This paper proposes an approach to solve binary combinatorial problems using continuous metaheuristics, focusing on the importance of binarization in the optimization process. Experimental results show that binarization rules have a greater impact on algorithm performance than transfer functions. It was found that sets of actions incorporating the elite or elite roulette binarization rule are the best.