Metaheuristic Algorithms

General-purpose optimization algorithms that can be applied to a wide variety of problems.

Overview

Metaheuristic algorithms are high-level problem-independent algorithmic frameworks that provide guidelines for developing heuristic optimization algorithms. They balance exploration and exploitation to efficiently search large solution spaces.

Available Algorithms

• Harmony Search - Music-inspired optimization
• Cross Entropy Method - Adaptive importance sampling
• Sine Cosine Algorithm - Mathematical function-based search
• Simulated Annealing - Thermodynamics-inspired (also in Classical)
• Tabu Search - Memory-based search
• Variable Neighborhood Search - Local search strategy

Usage Example

from opt.metaheuristic import HarmonySearch
from opt.benchmark.functions import rosenbrock

optimizer = HarmonySearch(
    func=rosenbrock,
    lower_bound=-5,
    upper_bound=10,
    dim=10,
    max_iter=100
)
best_solution, best_fitness = optimizer.search()

See Also

• API Reference - Complete API documentation
• Benchmark Results - Performance comparisons