BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Causality - Dr Ricardo Silva\, Gatsby Unit\, UCL
DTSTART:20061109T160000Z
DTEND:20061109T180000Z
UID:TALK5541@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:There is a difference between "seeing" and "doing". Consider t
 he following example: while one might observe that people in Florida live 
 longer\, this does not mean one should consider moving there to achieve lo
 ngevity. It is actually the case that many retired Americans move to Flori
 da\, which explains its high life expectancy. This illustrates that we can
  have a feature that is both useful for predicting life expectancy\, but a
 t the same time worthless when considered as a treatment.\n\nTo predict ef
 fects of interventions such as medical treatments\, public policies\, or e
 ven the behaviour of genetically engineered cells\, one needs a causal mod
 el. While there exists a well-established machinery designed to estimate s
 uch models using experimental data\, quite often one cannot perform experi
 ments for reasons such as high costs or ethical issues. Observational data
  (i.e.\, non-experimental)\, however\, can be easily collected in many cas
 es: data on the association between smoking and lung cancer is the classic
 al example.\n\nIn this talk we will explore several modern techniques of l
 earning causal effects from observational data. Such techniques rely on im
 portant assumptions linking statistical distributions to causal connection
 s and can be explored in many exciting ways by machine learning algorithms
 . Inferring causality from observational data is certainly among the harde
 st learning tasks of all\, but it is also a task with high pay-offs. \n\nT
 he outline of the talk is as follows:\n\n- Motivation and definitions / ob
 servational studies / the problems of directionality and confounding\n\n- 
 Languages for causal modeling: graphical models and potential outcomes\n\n
 - Identification of effects using graphical models\n\n- Notions of learnin
 g causal structure from data\n\n- Applications\n
LOCATION:LR4\, Engineering\, Department of
END:VEVENT
END:VCALENDAR