BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Semiparametric inference for impulse response functions using doub
 le/debiased machine learning
DTSTART:20260224T121500
DTEND:20260224T131500
DTSTAMP:20260415T164639Z
UID:01c60ef6ae348d095cc8194d18aaa16639c40f9e46468035a2053cc6
CATEGORIES:Conferences - Seminars
DESCRIPTION:Daniele Ballinari - Data Scientist at Swiss National Bank | Le
 cturer at University of St.Gallen\nWe introduce a double/debiased machine 
 learning estimator for the impulse response function in settings where a t
 ime series of interest is subjected to multiple discrete treatments\, assi
 gned over time\, which can have a causal effect on future outcomes. The pr
 oposed estimator can rely on fully nonparametric relations between treatme
 nt and outcome variables\, opening up the possibility to use flexible mach
 ine learning approaches to estimate impulse response functions. To this en
 d\, we extend the theory of double machine learning from an i.i.d. to a ti
 me series setting and show that the proposed estimator is consistent and a
 symptotically normally distributed at the parametric rate\, allowing for s
 emiparametric inference for dynamic effects in a time series setting. The 
 properties of the estimator are validated numerically in finite samples by
  applying it to learn the impulse response function in the presence of ser
 ial dependence in both the confounder and observation innovation processes
 . We also illustrate the methodology empirically by applying it to the est
 imation of the effects of  macroeconomic shocks.\n\nPaper
LOCATION:UNIL\, Extranef\, room 126
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR