Dollmaker Optimization Algorithm: A Novel Human-Inspired Optimizer for Solving Optimization Problems

International journal of intelligent engineering and systems, Journal Year: 2024, Volume and Issue: 17(3), P. 816 - 828

Published: May 3, 2024

In this article, a new human-based metaheuristic algorithm named Dollmaker Optimization Algorithm (DOA) is introduced, which imitates the strategy and skill of dollmaker when making dolls.The basic inspiration DOA derived from two natural behaviors in doll process (i) general changes to dollmaking materials (ii) precise small on appearance characteristics theory proposed then modeled mathematically phases exploration based simulation large made doll-making exploitation performance optimization evaluated twenty-three standard benchmark functions unimodal, high-dimensional multimodal, fixed-dimensional multimodal types.The results show that has achieved suitable for problems with its ability exploration, exploitation, balance them during search process.Comparison twelve competing algorithms shows superior compared by providing better all getting rank first best optimizer.In addition, efficiency handling real-world applications four engineering design problems.Simulation acceptable real world values variables objective algorithms.

Language: Английский

---

Enhanced Wombat Optimization Algorithm for Multi-Objective Optimal Power Flow in Renewable Energy and Electric Vehicle Integrated Systems

Results in Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 103671 - 103671

Published: Dec. 1, 2024

Language: Английский

---

Leveraging variational autoencoders and recurrent neural networks for demand forecasting in supply chain management: A case study

Journal of Infrastructure Policy and Development, Journal Year: 2024, Volume and Issue: 8(8), P. 6639 - 6639

Published: Aug. 26, 2024

Accurate demand forecasting is key for companies to optimize inventory management and satisfy customer efficiently. This paper aims Investigate on the application of generative AI models in forecasting. Two were used: Long Short-Term Memory (LSTM) networks Variational Autoencoder (VAE), results compared select optimal model terms performance accuracy. The difference actual predicted values also ascertain LSTM’s ability identify latent features basic trends data. Further, some research works focused computational efficiency scalability proposed methods providing guidelines implementation complicated techniques Based these results, LSTM have a promising enhancing consequently helpful decision-making process regarding control other resource allocation.

Language: Английский

---

SGA-Driven feature selection and random forest classification for enhanced breast cancer diagnosis: A comparative study

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 30, 2025

In this study, we propose a novel approach for breast cancer classification that integrates the Seagull Optimization Algorithm (SGA) feature selection with Random Forest (RF) classifier effective data classification. The novelty of our lies in first-time application SGA gene diagnosis, where systematically explores space to identify most informative subsets, thereby improving accuracy and reducing computational complexity. selected features are subsequently classified using RF, known its robustness high handling complex datasets. To evaluate effectiveness proposed method, compared it other classifiers, including Linear Regression (LR), Support Vector Machine (SVM), K-Nearest Neighbors (KNN). SGA-RF combination achieved best mean 99.01% 22 genes, outperforming methods demonstrating consistent performance across varying subsets. accuracies ranged from 85.35 94.33%, highlighting balance between reduction accuracy. Future work will explore integration nature-inspired algorithms deep learning models further enhance clinical applicability.

Language: Английский

---

Painting Training Based Optimization: A New Human-based Metaheuristic Algorithm for Solving Engineering Optimization Problems

Engineering Technology & Applied Science Research, Journal Year: 2025, Volume and Issue: 15(2), P. 21774 - 21782

Published: April 3, 2025

This study introduces a completely different perspective on optimization through the development of novel human-based metaheuristic algorithm named Painting Training Based Optimization (PTBO). Inspired by intricate and iterative human activities observed during painting training, PTBO models these creative systematic processes to effectively address challenges. The algorithm's foundation is rooted in concepts exploration exploitation, which are essential for achieving balance between searching solution space widely refining promising areas. theoretical framework comprehensively described, followed detailed mathematical modeling its two-phase operation. To evaluate capability, tested 22 constrained problems sourced from well-regarded CEC 2011 test suite. experimental results show that excels at producing competitive high-quality solutions. A comparative analysis with 12 other well-known algorithms underscores PTBO's superior performance, particularly handling complex benchmark functions. proposed approach outperformed competing all (22) findings highlight effectiveness solving real-world problems, showcasing potential outperform existing methods. By offering approach, contributes significant innovative tool challenges engineering applied domains.

Language: Английский

---

ICSO: A Novel Hybrid Evolutionary Approach with Crisscross and Perturbation Mechanisms for Optimizing Generative Adversarial Network Latent Space

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(10), P. 5228 - 5228

Published: May 8, 2025

Hybrid evolutionary approaches have gained significant attention for solving complex optimization problems, but their potential optimizing the low-dimensional latent space of generative adversarial networks (GANs) remains underexplored. This paper proposes a novel improved crisscross (ICSO) algorithm, hybrid approach that integrates and perturbation mechanisms to find suitable vector. The ICSO algorithm treats quality diversity as separate objectives, balancing them through normalization strategy, while gradient regularization term (i.e., GP) is introduced into discriminator’s objective function stabilize training mitigate gradient-related issues. By combining global local search capabilities particle swarm (PSO) with rapid convergence optimization, efficiently explores exploits space. extensive experiments demonstrate outperforms state-of-the-art algorithms in various classical GANs across multiple datasets. Furthermore, practical applicability validated its integration StyleGAN3 generating unmanned aerial vehicle (UAV) images, showcasing effectiveness real-world engineering applications. work not only advances field GAN also provides robust framework applying modeling tasks.

Language: Английский

---

An enhanced Pied kingfisher optimizer for UAV path planning and engineering design problems

Published: May 13, 2025

The Pied Kingfisher Optimizer (PKO) is an advanced optimization algorithm. Its slow convergence and propensity to become stuck in local optima are its drawbacks, though. We suggest Enhanced algorithm (EPKO) overcome these drawbacks. In order enhance the algorithm's exploratory position modifications make it easier identify global optimum, tent mapping adaptive T-distribution control approach used. Additionally, we present a Cauchy mutation method, which gives individuals strong ability avoid extrema guide population more advantageous directions. improve optimizer's search performance greatly boost accuracy, speed, stability for solving complicated issues, leader-based boundary technique also suggested. compare EPKO's against eight well-known algorithms number of dimensions using 29 CEC2017 benchmark functions. efficacy EPKO was demonstrated by fact that our came out on top every comparison. mathematically modeled UAV used variety competitor address path planning problem assess suggested method's practicality. tackled three engineering design challenges several methods. results show has best performance. When comes solution quality stability, generally performs better than competitors, demonstrating greater application potential.

Language: Английский

---

Salp Swarm Algorithm Optimized A* Algorithm and Improved B-Spline Interpolation in Path Planning

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(10), P. 5583 - 5583

Published: May 16, 2025

The efficiency and smoothness of path planning algorithms are critical factors influencing their practical applications. A traditional A* algorithm suffers from limitations in search efficiency, smoothness, obstacle avoidance. To address these challenges, this paper introduces an improved that integrates the Salp Swarm Algorithm (SSA) for heuristic function optimization proposes a refined B-spline interpolation method smoothing. first major improvement involves enhancing by optimizing its through SSA. combines Chebyshev distance, Euclidean density, with SSA adjusting weight parameters to maximize efficiency. simulation experimental results demonstrate modification reduces number searched nodes more than 78.2% decreases time over 48.1% compared algorithms. second key contribution is incorporating two-stage strategy smoother safer paths. corner avoidance adjusts control points near sharp turns prevent collisions, followed fine-tunes point positions ensure safe distances obstacles. show optimized increases minimum distance 0.2–0.5 units, improves average 43.0%, curvature approximately 61.8%. Comparative evaluations across diverse environments confirm superiority proposed computational safety. This study presents effective robust solution complex scenarios.

Language: Английский

---

A novel quality of service-aware task scheduling approach for cloud computing using modified Wombat Optimization Algorithm

Journal of Engineering and Applied Science, Journal Year: 2025, Volume and Issue: 72(1)

Published: May 26, 2025

Abstract This paper proposes a novel QoS-aware task scheduling approach in cloud computing environments that utilizes the Modified Wombat Optimization Algorithm (MWOA). Task is critical challenge computing. In Internet of Things (IoT) applications, there significant need to identify efficient methods for allocating computational resources or reducing cost and latency while maintaining reliability services. Most existing algorithms encounter difficulties terms premature convergence suboptimal performance multi-objective scenarios. MWOA addresses this by incorporating Levy flight enhanced global exploration chaotic sine map local exploitation, thus achieving delicate balance between speed solution accuracy. The key QoS factors optimized study include completion time, execution cost, resource consumption. Simulation results show reduces time 31% 17% compared traditional algorithms. work, we capabilities future development task-scheduling model toward unifying dynamic, computationally intensive cloud, heterogeneous, real-time IoT environments.

Language: Английский

---

Rabbit algorithm for global optimization

Applied Mathematical Modelling, Journal Year: 2024, Volume and Issue: unknown, P. 115860 - 115860

Published: Dec. 1, 2024

Language: Английский

---