BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Daisy - a framework for sound accuracy analysis and optimization o
 f numerical programs
DTSTART:20181130T141500
DTEND:20181130T151500
DTSTAMP:20260407T115056Z
UID:a1dd700311ca26fe390afb55349dd8e952cd0e9937ecd005fc5b238d
CATEGORIES:Conferences - Seminars
DESCRIPTION:By Eva Darulova\nVideo conference\n\nAbstract\nComputing resou
 rces are fundamentally limited and sometimes an exact solution may not eve
 n exist. Thus\, when implementing real-world systems\, approximations are 
 inevitable\, as are the errors they introduce. The magnitude of errors is 
 problem-dependent but higher accuracy generally comes at a cost in terms o
 f memory\, energy or runtime\, effectively creating an accuracy-efficiency
  tradeoff. To take advantage of this tradeoff\, we need to ensure that the
  computed results are sufficiently accurate\, otherwise we risk disastrous
 ly incorrect results or system failures. Unfortunately\, the current way o
 f programming with approximations is mostly manual\, and consequently cost
 ly\, error prone and often produces suboptimal results.\n\nIn this talk\, 
 we present the current state of the tool Daisy which approximates numerica
 l programs in an automated and trustworthy fashion. Daisy allows a program
 mer to write exact high-level code and generates an efficient implementati
 on satisfying a given accuracy specification. We discuss Daisy's verificat
 ion techniques for bounding the effects of numerical errors\, and the appr
 oximations Daisy can synthesize fully automatically.\n\nBio\nEva Darulova 
 obtained her PhD from EPFL in 2014. Since 2015\, she is a tenure-track fac
 ulty at the Max Planck Institute for Software Systems in Germany. Her rese
 arch interests include programming languages\, software verification and a
 pproximate computing. Recently\, she has also been involved with programmi
 ng robots.\n\nMore information\n\n 
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR