The folder contains a notebook on Bayesian optimization : how to tune a computer code simulating a catapult such that the projectile is thrown as far as possible. That kind of problem can be found in machine learning as well, where several tuning parameters must be found in order to minimize a cross-validation error.