<aside> 💡 A detailed overview incl. code examples can be found in The Kaggle Book p. 428 of 826 (56%)

</aside>

Scikit-optimize is built on top of Scikit-learn, and it extends its functionality by supporting Sequential model-based optimization in Python, that is, Bayesian Optimization. Through the class BayesianSearchCV, we can set up Bayesian optimization using Gaussian processes or Random forests to model the objective function, just as we would do with the GridSearchCV and RandomizedSearchCV from Scikit-learn.

Scikit-optimize also provides functions out-of-the-box to produce visualizations of the hyperparameter search, and better understand how the different hyperparameter values affect the model performance.