BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260627T211741EDT-2456vVdNvH@132.216.98.100
DTSTAMP:20260628T011741Z
DESCRIPTION:'Automatic Debiased Machine Learning via Neural Nets for Genera
 lized Linear Regression'\n\n>Whitney Newey (MIT)\n	Montreal Econometrics Se
 minar\n	September 30\, 2022\, 3:30 to 5:00 PM\n	ARTS 160\n\nAbstract:\n	We gi
 ve debiased machine learners of parameters of interest that depend on gene
 ralized linear regressions\, which regressions make a residual orthogonal 
 to regressors. The parameters of interest include many causal and policy e
 ffects. We give neural net learners of the bias correction that are automa
 tic in only depending on the object of interest and the regression residua
 l. Convergence rates are given for these neural nets and for more general 
 learners of the bias correction. We also give conditions for asymptotic no
 rmality and consistent asymptotic variance estimation of the learner of th
 e object of interest. We find that the resulting estimator of the average 
 treatment effect outperforms a state of the art neural net estimator based
  on inverse propensity score weighting in a simulation study.\n
DTSTART:20220930T193000Z
DTEND:20220930T210000Z
LOCATION:Room 160\, Arts Building\, CA\, QC\, Montreal\, H3A 0G5\, 853 rue 
 Sherbrooke Ouest
SUMMARY:Whitney Newey (MIT)\, 'Automatic Debiased Machine Learning via Neur
 al Nets for Generalized Linear Regression' 
URL:https://www.mcgill.ca/economics/channels/event/whitney-newey-mit-automa
 tic-debiased-machine-learning-neural-nets-generalized-linear-regression-34
 1758
END:VEVENT
END:VCALENDAR