Teaching data science by history: Kepler's laws of planetary motion and generalized linear models

Abstract

Teaching data science is challenging: it is a multidisciplinary subject that requires solid mathematical background. There are many models and approaches to consider. It is important, in our view, to present a unified approach to teaching this subject. We believe that one of the most e ective ways to do so is to present historical examples. An interesting historical example that explains Generalized Linear Models in prediction is the quest by the German astronomer, Johann Kepler, at the beginning of the 17-th century to find a unifying law explaining the motion of the planets in our Solar system.