BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Toward Causal Machine Learning - Prof. Dr. Bernhard Schölkopf - M
 ax Planck Institute for Intelligent Systems
DTSTART:20150429T130000Z
DTEND:20150429T140000Z
UID:TALK58956@talks.cam.ac.uk
CONTACT:David Greaves
DESCRIPTION:In machine learning\, we use data to automatically find depend
 ences in the world\, with the goal of predicting future observations. Most
  machine learning methods build on statistics\, but one can also try to go
  beyond this\, assaying causal structures underlying statistical dependenc
 es. Can such causal knowledge help prediction in machine learning tasks? W
 e argue that this is indeed the case\, due to the fact that causal models 
 are more robust to changes that occur in real world datasets. We touch upo
 n the implications of causal models for machine learning tasks such as dom
 ain adaptation\, transfer learning\, and semi-supervised learning. We also
  present an application to the removal of systematic errors for the purpos
 e of exoplanet detection.\n\nMachine learning currently mainly focuses on 
 relatively well-studied statistical methods. Some of the causal problems a
 re conceptually harder\, however\, the causal point of view can provide ad
 ditional insights that have substantial potential for data analysis.\n\n<i
 >Note</i>: This talk is also being given at MSR.
LOCATION:Lecture Theatre 1\, Computer Laboratory
END:VEVENT
END:VCALENDAR