Physics teaching mediated by Machine Learning: The case of the simple pendulum

In an era where the prevalence of data is striking, the importance of adopting advanced computational techniques, such as artificial intelligence and machine learning, becomes fundamental in the training process of tomorrow’s physicists. This article introduces an innovative approach to physics education, incorporating machine learning into the classic study of the simple pendulum. Key concepts of machine learning techniques are introduced and a broad dataset is collected through accurate observation of the pendulum’s motion under different initial conditions. From these data, a regression algorithm is employed to create a model of the pendulum system with satisfactory statistical indices. This pedagogical model contrasts with traditional physics teaching methods and opens a new horizon for understandding what it means to teach physics in the information age.

Keywords:
Machine Learning; Simple Pendulum; Physics Education