Computer Science Colloquium

Wolfgang Waltenberger
HEPHY -Institut für Hochenergiephysik, Wien

On Particle Physics, Information, and Machines That Learn

Thu 25.01.2018, 10:30, 60 minutes
Johannes Kepler University Linz, Room S2 219

Abstract

Recent advances in machine learning have not only innovated much of the hi-tech industry, they also change how modern science is pursued. By giving a few subjectively selected highlights from the field of particle physics, I wish to describe the data challenges that particle physics is currently facing, the solutions that have worked in the past, and possible ideas for the future. I shall however argue that not only does machine learning affect particle physics, physics might also help elucidate on machine learning, similar to how biology or neuroscience inspired many machine learning algorithms. "Why is deep learning so cheap? Can we apply the mathematics of curved spacetimes to information spaces? Could the notion of quantum mechanical superpositions help in developing efficient algorithms? Can machine learning benefit from quantum computing?" These are questions that physicists are currently debating. I shall briefly (and only superficially) touch upon these topics and some of their possible answers.

Bio

Dr Wolfgang Waltenberger Geboren 1974 in Gmunden, OÖ
Diplom 2001 in theoretischer Physik, TU Wien
Dissertation 2004 in Teilchenphysik / Statistik, TU Wien
Seit 2001 wissenschaftlicher Mitarbeiter der österr. Akademie der Wissenschaften
Seit 2001 Mitglied der CMS Kollaboration am CERN, Genf
Seit 2015 Vorlesungen in Astroteilchenphysik und Statistik an Uni Wien und TU Wien
Arbeitsschwerpunkte: Supersymmetrie, Interpretation der LHC Daten, Statistische Datenauswertung, Maschinelles Lernen
http://www.hephy.at/nc/de/das-hephy/mitarbeiterinnen/detail/name/waltenberger/

Invited by Prof. Dr. Sepp Hochreiter

---

The Computer Science Colloquium is organized by the Department of Coputer Science at JKU, the Österreichische Gesellschaft für Informatik (ÖGI) and the Österreichische Computergesellschaft (OCG).

List of all talks