Skip to content

Gradient-Free-Optimizers

The GFO backend is Hyperactive's primary optimization engine with 20+ algorithms spanning the complete spectrum of optimization methods. This is your go-to choice for maximum algorithmic flexibility and research applications.

Why Choose GFO?

  • Largest Algorithm Collection: 20+ methods from simple hill climbing to advanced Bayesian optimization
  • Full Customization: Direct access to all algorithm parameters and settings
  • Research-Friendly: Perfect for experimenting with different optimization strategies

Usage Pattern

Combine a GFO optimizer with an experiment and call solve():

from hyperactive.experiment.integrations import SklearnCvExperiment
from hyperactive.opt.gfo import HillClimbing

exp = SklearnCvExperiment(...)
opt = HillClimbing(experiment=exp)
best = opt.solve()

See the Quick Start for end‑to‑end examples and the Experiments section for building custom problems.