A review of artificial fish swarm algorithms: recent advances and applications

Farhad Pourpanah, Ran Wang, Chee Peng Lim, Xi-Zhao Wang, Danial Yazdani

TL;DR

This paper provides a comprehensive review of the Artificial Fish Swarm Algorithm (AFSA), a bio-inspired optimization technique. It surveys recent advancements, modifications, and applications of AFSA, highlighting its strengths and suggesting future research directions for this increasingly popular Swarm Intelligence methodology.

What Problem Does This Paper Address?

The paper addresses the need for a consolidated understanding of the Artificial Fish Swarm Algorithm (AFSA) and its evolution. AFSA, inspired by fish schooling behavior, has emerged as a flexible and effective Swarm Intelligence (SI) method for solving complex optimization problems. However, since its introduction, numerous improvements and hybrid models have been developed, making it challenging to keep track of the latest advancements and their applications. This review aims to bridge that gap by providing a structured overview of AFSA's progress, particularly focusing on continuous optimization problems from 2013 onwards.

Methods

The paper reviews AFSA through a taxonomic lens, categorizing modifications into:

  • Parameter Control: Adjusting parameters like visual distance (v), step length (s), and crowding factor (δ) to balance exploration and exploitation. This includes adaptive tuning methods.
  • Procedure and Sub-Functions: Enhancing or introducing new behaviors (e.g., leaping, swallow) beyond the original preying, swarming, and following.
  • Hybridization: Combining AFSA with other algorithms, categorized into:
  • Population-based methods: Merging AFSA with algorithms like Particle Swarm Optimization (PSO), Differential Evolution (DE), etc., either by integrating their dynamics or having them co-operate.
  • Non-population-based methods: Integrating AFSA with techniques like Fuzzy C-means (FCM) or Support Vector Regression (SVR).
  • Multi-objective AFSA: Adapting AFSA to solve problems with multiple conflicting objectives.
  • Dynamic AFSA: Modifying AFSA to handle optimization problems where the environment changes over time.

The review focuses on articles published in key indexed journals and analyzes modifications to parameters, procedures, and sub-functions, along with their associated applications.

Datasets

Not explicitly discussed as the paper is a review and survey of existing literature. The authors mention using benchmark functions for empirical studies, but specific datasets for training or evaluation of AFSA variants are not detailed.

Key Insights and Ideas

  • AFSA's Versatility: AFSA's core behaviors (preying, swarming, following) provide a robust foundation for tackling a wide range of continuous, binary, and combinatorial optimization problems.
  • Balancing Exploration and Exploitation: A significant portion of AFSA modifications focuses on fine-tuning parameters to achieve an optimal balance between exploring new areas of the search space and exploiting promising regions.
  • Hybridization is Key: Combining AFSA with other SI algorithms or machine learning techniques often yields superior performance by leveraging the strengths of each component.
  • Addressing Dynamic Environments: Dynamic AFSA models are crucial for real-world scenarios where problem landscapes change, requiring algorithms that can not only find optima but also track them.
  • Maintaining Diversity: A key challenge and area of research is preventing AFSA from premature convergence by maintaining population diversity, especially in complex or multi-modal problems.

Key Results

The paper does not present new experimental results but synthesizes findings from over 1000 reviewed articles. Key qualitative findings include:

  • AFSA has shown good performance compared to other population-based optimization models.
  • Hybrid AFSA models often outperform standalone AFSA and other algorithms.
  • Adaptive parameter tuning methods have been successful in improving AFSA's convergence speed and robustness.
  • The publication trend of AFSA shows a consistent upward trajectory, indicating its growing importance.

Limitations and Future Work

Limitations:

  • Computational Complexity: The original AFSA can be computationally intensive due to distance calculations.
  • Convergence Speed: AFSA's convergence speed can be slow, especially in later stages, due to movement limitations.
  • Theoretical Underpinnings: Empirical studies dominate AFSA research, with less focus on theoretical analysis of its convergence properties and runtime efficiency.
  • Search Space Exploration: The effectiveness of AFSA in highly complex or rotated search spaces with strong variable interactions is still an area needing more investigation.

Future Work:

  • Improving Computational Speed: Developing more efficient AFSA variants.
  • Enhancing Convergence Speed: Exploring new strategies to accelerate convergence without sacrificing exploration.
  • Theoretical Analysis: Conducting principled theoretical studies on AFSA's convergence, complexity, and behavior in different search spaces.
  • Addressing Large-Scale and Multi-modal Problems: Further adapting AFSA for high-dimensional problems and those with multiple optima.
  • Dynamic Optimization: Developing more robust AFSA models for dynamic environments, focusing on diversity maintenance and tracking capabilities.
  • Discrete AFSA Models: The authors indicate their future work will focus on a comprehensive review of discrete AFSA models.

Summary updated

Abstract

The Artificial Fish Swarm Algorithm (AFSA) is inspired by the ecological behaviors of fish schooling in nature, viz., the preying, swarming and following behaviors. Owing to a number of salient properties, which include flexibility, fast convergence, and insensitivity to the initial parameter settings, the family of AFSA has emerged as an effective Swarm Intelligence (SI) methodology that has been widely applied to solve real-world optimization problems. Since its introduction in 2002, many improved and hybrid AFSA models have been developed to tackle continuous, binary, and combinatorial optimization problems. This paper aims to present a concise review of the continuous AFSA, encompassing the original ASFA, its improvements and hybrid models, as well as their associated applications. We focus on articles published in high-quality journals since 2013. Our review provides insights into AFSA parameters modifications, procedures and sub-functions. The main reasons for these enhancements and the comparison results with other hybrid methods are discussed. In addition, hybrid, multi-objective and dynamic AFSA models that have been proposed to solve continuous optimization problems are elucidated. We also analyse possible AFSA enhancements and highlight future research directions for advancing AFSA-based models.
BibTeX
@article{pourpanah2023review,
  author = {Farhad Pourpanah and Ran Wang and Chee Peng Lim and Xi-Zhao Wang and Danial Yazdani},
  title = {{A review of artificial fish swarm algorithms: recent advances and applications}},
  journal = {Artif. Intell. Rev.},
  year = {2023},
  doi = {10.1007/s10462-022-10214-4},
}
Figure