Skip to content

Hyperparameter Optimization

KiteML includes hyperparameter search strategies to fine-tune candidate models.

1. Search Strategies

  • Randomized Search: Samples hyperparameter distributions efficiently.
  • Optuna Integration: Uses Bayesian optimization algorithms to maximize metric performance.
from kiteml.optimization import HyperparameterTuner

tuner = HyperparameterTuner(strategy="optuna", n_trials=50)
best_params = tuner.tune(model, X, y)