Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

AGPhil: Arbeitsgruppe Philosophie der Physik

AGPhil 6: Symposium: Epistemology of Big Data in Physics IV

AGPhil 6.2: Hauptvortrag

Freitag, 17. März 2017, 11:00–11:45, GW2 B2900

Causation, probability and all that: Data science as a novel inductive paradigm — •Wolfgang Pietsch — Munich Center for Technology in Society, TU München, Germany

Some have claimed that genuine data science is impossible since inductivism has allegedly long been refuted as a sound scientific methodology. Instead, I argue that data science stands in an old and venerable empiricist tradition which includes highly influential scientists and methodologists like Francis Bacon, Isaac Newton or John Stuart Mill. An inductive methodology is sketched that could serve as a fundamental conceptual framework for data science. On this basis, I disentangle the conceptual muddle behind the claim that correlation replaces causation in data science, which is often held to be one of the central tenets in this discipline. By contrast, causation remains the crucial concept to distinguish between meaningful and accidental correlations, i.e. between those that allow for prediction and manipulation and those that do not.