BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:BIG Hype: Best Intervention in Games via Distributed Hypergradient
  Descent
DTSTART:20251031T140000
DTEND:20251031T150000
DTSTAMP:20260917T083810Z
UID:5bf58fcbc9dec806151f7d8a3cfaf1f995ba9a74ddc3240cc783903f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Mr. Panagiotis Grontas\, PhD Automatic Control Laboratory (IfA
 )\, ETH Zurich.\nAbstract: \nHierarchical decision making problems\, such
  as bilevel programs and Stackelberg games\, are attracting increasing int
 erest in both the engineering and machine learning communities. Yet\, exis
 ting solution methods lack either convergence guarantees or computational 
 efficiency\, due to the absence of smoothness and convexity. In this work\
 , we bridge this gap by designing BIG Hype\, a first-order hypergradient-b
 ased algorithm for Stackelberg games\, and mathematically establishing its
  convergence using tools from nonsmooth analysis. To evaluate the hypergra
 dient\, namely\, the gradient of the upper-level objective\, we develop an
  online scheme that simultaneously computes the lower-level equilibrium an
 d its Jacobian. We demonstrate BIG Hype’s potential by deploying it on v
 arious large-scale bilevel problems\, such as demand response\, traffic ro
 uting\, and recommender systems.\n\nBio sketch:\nPanagiotis Grontas is a d
 octoral student in the Automatic Control Laboratory (IfA) at ETH Zürich\,
  supervised by John Lygeros. He received an M.Sc. degree with distinction 
 in Robotics\, Systems\, and Controls from ETH Zürich\, and a Diploma degr
 ee in Mechanical Engineering from the National Technical University of Ath
 ens. He is a recipient of the ETH Medal for outstanding Master’s theses 
 and the IEEE CSS Swiss Chapter Young Author Best Journal Paper Award 2025.
  His research is focused on the broad intersection of optimization\, machi
 ne learning\, and control.\n 
LOCATION:ME C2 405 https://plan.epfl.ch/?room==ME%20C2%20405 https://epfl.
 zoom.us/j/63001882786?pwd=G5ulqpMfWHZC8fbnFZwJf6uKI9hGBa.1
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR