Teaching Philosophy of Science to Data Scientists

Abstract

This article advocates for the integration of a Philosophy of Science course directed to future data scientists. Following similar courses that were often found in Physics or Biology study programs, the course focuses on the critical examination of the foundational principles underlying data science. This article highlights the need for data scientists to question the often-undisputed assumptions guiding their decision-making processes. The course developed and taught successfully to Master of Data Science students at the University of Luxembourg, delves into philosophical issues such as Hume's problem of induction, linking them with contemporary challenges in artificial intelligence and healthcare. From exploring scientific explanation and causation to addressing biases, confounders, and ethical considerations, the curriculum bridges theoretical concepts with practical applications, all of it illustrated with historical examples. The lessons from Philosophy of Science give the students essential knowledge to face the evolving technological environment and face current and future challenges in data science. This work describes the contents of the course and discusses the impact of these teachings, including the students’ opinions.

Ähnliche Arbeiten