The aim of machine learning is to make a computer learn from data without explicitly programming it to do so. The fruits of machine learning are all around us: email spam filters classify your messages using machine learning techniques, and postal services around the world "read" and sort hand-written addresses using machine learning techniques. The rise of machine learning upends traditional assumptions about hard problems in epistemology -- Hume's problem, Duhem's problem, and much of traditional analytic epistemology -- as well as traditional presumptions of expertise in statistics concerning the selection of models for data and the setting of a model's parameters. But the success of these methods also raises serious ethical and public policy questions.
This course is an introduction to this philosophically rich and radical method of inquiry. As logic was to 20th century philosophy, so computational methods shall be to this century. So, the course is designed to be a hands-on introduction to machine learning with an emphasis on its philosophical underpinnings and methodological implications.