HPSO-LS is a feature subset selection algorithm that combines Hybrid Particle Swarm Optimization (PSO) with a Local Search strategy. This repository provides the implementation of the HPSO-LS algorithm.
- Integration of PSO and Local Search for efficient feature subset selection.
- Customizable algorithm parameters to adapt to different optimization goals.
- Compatibility with popular machine learning frameworks and libraries.
- Visualization and analysis of results for better understanding and evaluation.
- Comprehensive documentation and code examples for easy implementation.
The HPSO-LS algorithm uses the "Sonar, Mines vs. Rocks" dataset for demonstration purposes. The dataset is included in the repository and can be found in the sonar.mines and sonar.rocks files. It consists of sonar signals and aims to discriminate between metal cylinders and rocks.
The HPSO-LS algorithm provides various parameters that can be customized:
- Swarm size: Number of particles in the swarm.
- Maximum iterations: Maximum number of iterations for the optimization process.
- Cognitive coefficient: Weight for the particle's best-known position.
- Social coefficient: Weight for the global best-known position.
- Inertia weight: Controls the impact of the particle's previous velocity.
- Local search iterations: Number of iterations for the local search phase.
The HPSO-LS algorithm aims to find the optimal feature subset that maximizes the performance of machine learning models. The optimization goals include:
- Maximizing classification accuracy.
- Minimizing training time.
- Reducing overfitting by selecting the most relevant features.
For any questions or inquiries, please contact Rafe Sharif at [email protected].