BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:STATISTICAL INFERENCE IN LARGE MULTI-WAY NETWORKS
DTSTART:20260615T090000Z
DTEND:20260615T100000Z
DTSTAMP:20260614T131300Z
UID:indico-event-16426@indico.math.cnrs.fr
DESCRIPTION:Speakers: Lucas Resende (Postdoc CREST)\n\nWe propose a new me
 thod to estimate structural parameters in multi-way networks while control
 ling for rich structures of fixed effects. The method is based on a serie
 s of classification tasks and is agnostic to both the number and structure
  of fixed effects. In contrast to full maximum likelihood approaches\, our
   estimator  does not suffer from the incidental parameter problem. For 
 sparsely connected networks\, it is also computationally faster than PPML.
  We provide empirical evidence that our estimator yields more reliable con
 fidence intervals than PPML and its bias-correction strategies. These impr
 ovements hold even under model misspecification and are more pronounced in
  sparse settings. While PPML remains competitive in dense\, low-dimensiona
 l data\, our approach offers a robust alternative for multi-way models tha
 t scales efficiently with sparsity. The method is applied to study the cau
 sal effect of a policy reform on spatial accessibility to health care in F
 rance.\n\nhttps://indico.math.cnrs.fr/event/16426/
LOCATION:Fokko du Cloux\, 1er étage bât Braconnier (La doua)
URL:https://indico.math.cnrs.fr/event/16426/
END:VEVENT
END:VCALENDAR