BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//juliacon-2022//speaker//ZME37L
BEGIN:VEVENT
SUMMARY:InferOpt.jl: combinatorial optimization in ML pipelines - Guillaum
 e Dalle\, Louis Bouvier\, Léo Baty
DTSTART:20220729T170000Z
DTEND:20220729T173000Z
DTSTAMP:20260817T083126Z
UID:pretalx-juliacon-2022-P7XJCV@pretalx.com
DESCRIPTION:We present InferOpt.jl\, a generic package for combining combi
 natorial optimization algorithms with machine learning models. It has two 
 purposes:\n\n- Increasing the expressivity of learning models thanks to ne
 w types of structured layers.\n- Increasing the efficiency of optimization
  algorithms thanks to an additional inference step.\n\nOur library provide
 s wrappers for several state-of-the-art methods in order to make them comp
 atible with Julia's automatic differentiation ecosystem.
LOCATION:Purple
URL:https://pretalx.com/juliacon-2022/talk/P7XJCV/
END:VEVENT
END:VCALENDAR