gradient-swarm-optimization

GSO: Gradient-guided Swarm Optimization

A Python implementation of the Gradient-guided Swarm Optimization algorithm, designed for solving binary optimization problems. The algorithm uses a mathematically grounded approach combining gradient-based directional guidance with neighborhood learning mechanisms.

Features

• Binary optimization for various problem types
• Gradient-guided search for effective exploration
• Neighborhood-based learning for local exploitation
• Built-in command line interface
• Automatic result saving and history tracking
• Early stopping with known optimum
• Automatic progress reporting
• Built-in timeout mechanism
• Reproducible results with seed setting

Installation

pip install gso

Quick Start

from gso import GSO
import numpy as np

def simple_fitness(solution):
    """Example fitness function: maximize sum of elements."""
    return np.sum(solution)

GSO.run(
    obj_func=simple_fitness,
    dim=100,
    pop_size=1000,
    max_iter=800,
    obj_type='max',
    neighbour_count=3,
    gradient_strength=0.8,
    base_learning_rate=0.1
)

Test Examples and Data

To run the test examples (UFLP and DUF problems):

• Clone the GitHub repository:

git clone https://github.com/gazioglue/gso.git
cd gso

• Run UFLP solver:

python examples/solve_uflp.py examples/data/uflp/test_instances/cap71.txt

• Run DUF solver:

python examples/solve_dufs.py duf1

Available Command Line Options

For UFLP:

python solve_uflp.py cap71.txt --pop-size 2000 --max-iter 1000 --seed 42

For DUF:

python solve_dufs.py duf2 --dim 200 --pop-size 2000 --seed 42

Documentation

For more detailed usage instructions and examples, see USAGE.md.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like change.

License

This project is licensed under the MIT License - see the LICENSE file for details.