Enhancing AI Optimization with Chaotic Maps: The Oscillating Chaotic Sunflower Optimization Algorithm

Abstract

Metaheuristic algorithms have been at the forefront of optimization research for many years, with continuous advancements and the development of new algorithms. Among these, the recently proposed Sunflower Optimization Algorithm (SFO) has emerged as a notable search algorithm due to its simplicity and effectiveness. However, as a relatively new algorithm, it presents opportunities for further enhancement and flexibility in its methodology. This study introduces the Oscillating Chaotic Sunflower Optimization Algorithm (OCSFO), an innovative variant of the SFO that incorporates a novel exploration technique utilizing chaotic maps. Specifically, the OCSFO algorithm employs Chebyshev, Circle, Logistic, Sine, and Tent chaotic maps to guide individual production and algorithm execution. The novelty of this research lies in the integration of chaotic dynamics into the SFO framework, enhancing its exploratory capabilities and potentially improving convergence rates and solution quality. The OCSFO algorithm was applied to solve various optimization problems, including benchmark test functions and practical applications such as parameter tuning in machine learning models and optimizing design parameters in engineering systems. To evaluate the performance of the proposed OCSFO, both restricted and unrestricted test functions were utilized, providing a comprehensive assessment of its effectiveness. Comparative results demonstrate that the OCSFO achieves competitive outcomes compared to the classical SFO, underscoring its potential as a robust optimization tool. This work is highly relevant as it contributes to the ongoing evolution of metaheuristic algorithms, offering a new approach to optimization that leverages the strengths of chaotic systems. The findings of this study provide valuable insights and pave the way for further research and development in the field of metaheuristic optimization and its applications in artificial intelligence and engineering.