Fish School Search: A Weight-Based Swarm Optimization Method
Summary
The article describes Fish School Search, a population-based metaheuristic for continuous optimization. Each candidate solution is represented as a fish with a weight reflecting search success. The method combines individual movement, a shared instinctive adjustment based on recent fitness changes, and collective movement toward or away from the school’s weighted center. Feeding updates fish weights, while the coordinated operators balance local exploration and movement of the group through the search space.
It provides the update logic and an MQL5 implementation, then compares FSS with other randomized optimizers on benchmark functions. The reported results show FSS finding useful solutions but not consistently leading the comparison; the discussion characterizes it as relatively capable on smooth functions while noting uncertainty for functions with many variables. As a stochastic heuristic, it does not guarantee an exact optimum. Performance depends on parameters and the objective landscape, and benchmark behavior alone does not demonstrate gains in trading applications.
Key ideas
- Fish School Search assigns each candidate a weight that records its relative success during optimization.
- Individual swimming explores locally, while instinctive movement combines fitness-weighted position changes across the school.
- Collective-volitional movement expands or contracts the search around the weighted center depending on overall progress.
- The method targets continuous optimization and can be adapted to other problem types.
- Benchmark comparisons show useful but inconsistent convergence, with no guarantee of a precise solution.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.