Classical Optimization Algorithms

Well-established optimization methods with proven convergence properties.

Overview

Classical optimization algorithms include direct search methods, quasi-Newton methods, and derivative-free techniques that have been studied extensively in optimization theory.

Available Algorithms

• BFGS - Broyden–Fletcher–Goldfarb–Shanno algorithm
• Nelder-Mead - Simplex-based derivative-free method
• Simulated Annealing - Thermodynamics-inspired probabilistic technique
• Hill Climbing - Local search algorithm
• Powell's Method - Conjugate direction method
• Trust Region - Constrained optimization approach

Usage Example

from opt.classical import NelderMead, SimulatedAnnealing
from opt.benchmark.functions import rosenbrock

# Nelder-Mead
nm = NelderMead(
    func=rosenbrock,
    lower_bound=-5,
    upper_bound=10,
    dim=10,
    max_iter=1000
)
best_solution, best_fitness = nm.search()

See Also

• API Reference - Complete API documentation
• Gradient-Based Methods - For gradient-based classical methods