Bayesian optimization is a powerful approach to hyperparameter tuning and optimization in machine learning, particularly effective for complex functions with unknown gradients and high evaluation costs. Unlike traditional methods like grid or random search, which often lack efficiency, Bayesian optimization combines probabilistic modeling and acquisition functions to balance exploration and exploitation, directing the search towards promising regions. This presentation explores the fundamentals of Bayesian optimization, highlighting its core components: the probabilistic model and the acquisition function. Through illustrative examples, we demonstrate Bayesian optimization’s applicability in tackling optimization challenges, offering an efficient strategy to minimize costly evaluations and identify optimal solutions.