A Survey on Swarm Intelligence Algorithms Employed for Optimizing Machine Learning Techniques Used in Recommendation Systems DOI

Lim Cher Zet,

Muhammed Basheer Jasser, Richard T.K. Wong

et al.

Published: Dec. 16, 2023

With the exponential growth of digital data, recommendation systems or recommender are widely used in various domains assisting users filtering and decision-making on massive information. Recommendation capable delivering personalized content to enhance user experience satisfaction through users' preferences behaviors. The machine learning algorithms employed facilitate effectiveness tasks achieved by those among which, for example, is providing accurate prediction that matches preferences. Swarm Intelligence offers robust optimization mechanisms have been successfully applied computational problems including refining algorithms. To best our knowledge, there no recent comprehensive survey swarm intelligence optimizing techniques when systems. Therefore, this presents a We conducted literature intelligence, using relevant keywords their variants, focusing publications since 2019. Our findings highlight use primarily clustering, classification, feature selection has significantly enhanced systems, especially clustering classification. However, balance between complexity processing speed remains challenge. Future research could focus these better efficiency

Language: Английский

Hybrid Beluga whale and jellyfish search optimizer for optimizing proton exchange membrane fuel cell parameter estimation DOI
Mohammad Aljaidi, Pradeep Jangir,

Arpita Arpita

et al.

Ionics, Journal Year: 2025, Volume and Issue: unknown

Published: April 14, 2025

Language: Английский

Citations

0
Utilizing the Honeybees Mating-Inspired Firefly Algorithm to Extract Parameters of the Wind Speed Weibull Model DOI Creative Commons
Abubaker Younis,

Fatima Belabbes,

Petru Adrian Cotfas

et al.

Forecasting, Journal Year: 2024, Volume and Issue: 6(2), P. 357 - 377

Published: May 22, 2024

This study introduces a novel adjustment to the firefly algorithm (FA) through integration of rare instances cannibalism among fireflies, culminating in development honeybee mating-based (HBMFA). The IEEE Congress on Evolutionary Computation (CEC) 2005 benchmark functions served as rigorous testing ground evaluate efficacy new diverse optimization scenarios. Moreover, thorough statistical analyses, including two-sample t-tests and fitness function evaluation analysis, algorithm’s capabilities were robustly validated. Additionally, coefficient determination, used an objective function, was utilized with real-world wind speed data from SR-25 station Brazil assess applicability modeling parameters. Notably, HBMFA achieved superior solution accuracy, enhancements averaging 0.025% compared conventional FA, despite moderate increase execution time approximately 18.74%. Furthermore, this dominance persisted when performance other common algorithms. However, some limitations exist, longer HBMFA, raising concerns about its practical scenarios where computational efficiency is critical. while demonstrates improvements values, establishing significance these differences FA not consistently achieved, which warrants further investigation. Nevertheless, added value work lies advancing state-of-the-art algorithms, particularly enhancing accuracy for critical engineering applications.

Language: Английский

Citations

3
A Kepler optimization algorithm improved using a novel Lévy-Normal mechanism for optimal parameters selection of proton exchange membrane fuel cells: A comparative study DOI Creative Commons
Mohamed Abdel‐Basset, Reda Mohamed, Karam M. Sallam

et al.

Energy Reports, Journal Year: 2024, Volume and Issue: 11, P. 6109 - 6125

Published: June 1, 2024

Proton exchange membrane fuel cells (PEMFCs) are considered a promising renewable energy source and have sparked lot of interest over the last few years due to their robust efficiency, low operating temperature, longevity. The PEMFC's electrochemical model has seven unknown parameters, which not given in manufacturer's datasheets need be accurately estimated present more accurate model, leading improved efficiency performance PEMFC systems. estimation those parameters been dealt with as complex non-linear optimization problem that needs powerful algorithm solve it. existing algorithms still some disadvantages, such falling into local minima convergence speed, make them ineligible this complicated acceptable accuracy computational cost. Therefore, study presents new parameter technique for estimating accurately, thereby achieving precise modeling PEMFCs. This called IKOA is based on integrating Kepler (KOA) novel Lévy-Normal (LN) mechanism strengthen its exploration exploitation capabilities against multimodal problem. Lévy flight aims improve KOA's operator accelerate speed toward near-optimal solution, thus minimizing cost; meanwhile, normal distribution used operator, aiding escape minima. proposed KOA herein evaluated several rival using six well-known commercial stacks highlight effectiveness. Key metrics cost, fitness measures, statistical validation through Wilcoxon rank-sum test IKOA's effective enhancing predictive operational numerical findings show high superiority all optimizers solved benchmarks.

Language: Английский

Citations

3
A Clustering Algorithm Employing Salp Swarm Algorithm and K-Means DOI
Bibi Aamirah Shafaa Emambocus, Muhammed Basheer Jasser,

Lim Cher Zet

et al.

Published: March 1, 2024

Clustering, one of the main types unsupervised machine learning, consists grouping data into clusters to discover hidden patterns. Hence it is a crucial learning task. The predominant algorithm employed for clustering tasks k-means algorithm. However, has some limitations including being sensitive initial centroids. Recently swarm intelligence algorithms have been noticed be able effectively optimize k-means. Hence, in this paper, Salp Swarm Algorithm (SSA), recent with favorable exploration and exploitation capabilities, optimizing Specifically, SSA centroids overcome its limitation. proposed applied as part movie recommendation system cluster users based on their preferences. experimental findings demonstrate that comparison original technique, yields superior outcomes clustered by lower within sum squares higher silhouette score.

Language: Английский

Citations

2
An Improved Nutcracker Optimization Algorithm for Discrete and Continuous Optimization Problems: Design, Comprehensive Analysis, and Engineering Applications DOI Creative Commons
Mohamed Abdel‐Basset, Reda Mohamed, Ibrahim M. Hezam

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(17), P. e36678 - e36678

Published: Aug. 23, 2024

This study is presented to examine the performance of a newly proposed metaheuristic algorithm within discrete and continuous search spaces. Therefore, multithresholding image segmentation problem parameter estimation both proton exchange membrane fuel cell (PEMFC) photovoltaic (PV) models, which have different spaces, are used test verify this algorithm. The traditional techniques could not find approximate solutions for those problems in reasonable amount time, so researchers algorithms overcome shortcomings. However, majority still suffer from slow convergence speed stagnation into local minima problems, makes them unsuitable tackling these optimization problems. proposes an improved nutcracker (INOA) better solving acceptable time. INOA based on improving standard using improvement strategy that aims improve prevent minima. first applied estimating unknown parameters single-diode, double-diode, triple-diode models PV module solar cell. Second, four PEMFC modules further observe INOA's challenge. Finally, investigated multi-thresholding its effectiveness space. Several images with threshold levels were validate effectiveness, stability, scalability. Comparison several rival optimizers various indicators, such as curve, deviation, average fitness value, Wilcoxon rank-sum test, demonstrates effective alternative Quantitively, solve than other optimizers, rates final results ranging between 0.8355 % 3.34 4.97 99.9

Language: Английский

Citations

1
A Two-Stage Genetic Algorithm for Beam–Slab Structure Optimization DOI Creative Commons
Zhexi Yang, Jane W. Z. Lu

Buildings, Journal Year: 2024, Volume and Issue: 14(9), P. 2932 - 2932

Published: Sept. 16, 2024

Beam–slab structures account for 50–65% of a building’s total dead load and contribute to 20% the overall cost CO2 emissions. Despite their importance, conventional beam–slab structural optimization methods often lack search efficiency accuracy, making them less effective practical engineering applications. Such limitations arise from problem involving complex solution space, particularly when considering components’ arrangement, dimensions, transfer paths simultaneously. To address research gap, this study proposes novel two-stage genetic algorithm, optimizing layout in first stage component topological relationships dimensions second stage. Numerical experiments on prototype case indicate that algorithm can generate results meet accuracy requirements within 100 iterations, outperforming comparable algorithms both accuracy. Additionally, heuristic approach stands out its independence prior dataset training minimal parameter adjustment requirement, it highly accessible engineers without programming expertise. Statistical analysis algorithm’s process studies demonstrate robustness adaptability various problems, revealing significant potential scenarios.

Language: Английский

Citations

1
Parameter Estimation of Three-Diode Photovoltaic Model Using Reinforced Learning-Based Parrot Optimizer with an Adaptive Secant Method DOI Open Access

Nandhini Kullampalayam Murugaiyan,

C. Kumar,

Magdalin Mary Devapitchai

et al.

Sustainability, Journal Year: 2024, Volume and Issue: 16(23), P. 10603 - 10603

Published: Dec. 3, 2024

In the developing landscape of photovoltaic (PV) technology, accuracy in simulating PV cell behaviour is dominant for enhancing energy conversion efficiency. This study introduces a new approach parameter estimation three-diode model, basis representation characteristics. The methodology combines reinforced learning-based parrot optimizer (RLPO) with an adaptive secant method (ASM) to fine-tune parameters governing model. RLPO algorithm inspired by mimetic ability parrots, i.e., foraging, staying, communicating, and fear noticed trained Pyrrhura Molinae as it influences learning mechanisms adaptively explore exploit search space optimal sets. Simultaneously, ASM enhances convergence rate through iterative adjustment mechanism, responding curvature objective function, thereby ensuring estimation. combination addresses complexities non-linearities inherent offering robust framework Through extensive simulations, proposed demonstrated superior performance terms accuracy, speed, reliability when compared existing algorithms. empirical results emphasize effectiveness integrating strategy handling details model parameterization. These outcomes show that can handle issues related optimization systems, opening door progress sustainable technologies.

Language: Английский

Citations

1
The evaluation of pressure swing thermal coupling technology based on intelligent optimization algorithms in the ethylbenzene/styrene separation process DOI
Hao Fu,

Xin Yang,

Yifeng Liu

et al.

Separation and Purification Technology, Journal Year: 2024, Volume and Issue: unknown, P. 130993 - 130993

Published: Dec. 1, 2024

Language: Английский

Citations

1
A brief survey of deep learning methods for android Malware detection DOI
Abdurraheem Joomye, Mee Hong Ling, Kok‐Lim Alvin Yau

et al.

International Journal of Systems Assurance Engineering and Management, Journal Year: 2024, Volume and Issue: 16(2), P. 711 - 733

Published: Dec. 20, 2024

Language: Английский

Citations

1
An Adaptive Chaotic League Championship Algorithm for Solving Global Optimization and Engineering Design Problems DOI

Jatsada Singthongchai,

Tanachapong Wangkhamhan

Published: Jan. 1, 2024

Language: Английский

Citations

0