Publication:
Alligator optimization algorithm: a novel bio-inspired optimization technique and its diverse applications

Loading...
Thumbnail Image
Date
2024-12-01
Authors
Tan, Weng Hooi
Journal Title
Journal ISSN
Volume Title
Publisher
Research Projects
Organizational Units
Journal Issue
Abstract
This research work proposes the Alligator Optimization (AgtrO) algorithm, a novel bio-inspired optimization algorithm that mimics the natural behaviors of alligators to enhance problem-solving in single-objective optimization tasks. It introduces several innovative methodologies to meet the research objectives. Through its dual-phase optimization approach, consisting of the hunting phase for global search and the relocating phase for local search, AgtrO strikes a powerful balance between exploration and exploitation, effectively addressing the exploration-exploitation dilemma. Novel concept of global exploitation and local exploration enhances AgtrO's ability to avoid and escape local optima, ensuring the consistent identification of global optimal solutions. Additionally, AgtrO incorporates novel adaptive parameters that enable it to handle scalability and dimensionality challenges in large datasets and complex problem domains. AgtrO was rigorously tested across multiple benchmark sets and real-world applications. The results substantiate AgtrO's ability to meet all predefined research objectives, mainly to assess whether it can achieve superior accuracy, precision and robustness while maintaining low computational complexity and time. In fact, the AgtrO algorithm has demonstrated exceptional performance across a wide range of benchmark sets and real-world applications, with all achieving accuracy, precision and robustness exceeding 91%. With an overall accuracy of 96.63%, an overall precision of 96.31%, and an overall robustness of 97.37%, AgtrO not only meets but surpasses the performance of recent advanced optimization algorithms. Its ability to minimize error, ensure consistency, and maintain stability under varying conditions highlights its reliability and adaptability. These results also support the claim that AgtrO is capable of maintaining strong performance even as the complexity and size of the problem increase, highlighting AgtrO's minimal issues with scalability and dimensional challenges. Moreover, its computational efficiency, with low execution times averaging 0.92 seconds, further solidifies its standing as an effective and competitive optimization algorithm. These results confirm AgtrO as a significant advancement in optimization, showcasing its potential for high-performance applications across diverse fields.
Description
Keywords
Citation