BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:On the Optimal Bias-Variance Trade-off in High Dimensions
DTSTART:20230515T110000
DTEND:20230515T120000
DTSTAMP:20260929T171652Z
UID:013d2d049cf2ccea12db62305fdd247699de30e04cf6f2618520b335
CATEGORIES:Conferences - Seminars
DESCRIPTION:Riccardo Cescon\, University of Padova\, Italy\nAbstract:\nIn 
 this talk we will concentrate on the classical bias-variance trade-off in
  linear regression\, for problems whose dimension and number of sample
 s are very large. Such problems present many challenges\, such as the hig
 h-dimensionality and the nonconvexity of the variance regularizer. We a
 re interested in studying the performance of the estimators and in unders
 tanding how to optimally choose the variance regularization parameter. We
  will show that\, despite the aforementioned challenges\, such problems c
 an be tackled using tools from distributionally robust optimization and 
 high-dimensional statistics.\n\nBio:\nRiccardo Cescon holds a bachelor’s
  degree in Information Engineering (2020) and a master’s degree in Contr
 ol Systems Engineering (2022) both with honors from the University of Pado
 va\, Italy. During his master’s thesis he spent six months at ETHZ as a 
 visiting student at the Institut für Automatik under the supervision of p
 rof. Florian Dörfler working on the optimal bias-variance tradeoff in hig
 h dimensions. He’s currently a research assistant at the Department of I
 nformation Engineering at the University of Padova under the supervision o
 f prof. Ruggero Carli on the European project Drapebot.\nHis research inte
 rests include control theory\, machine learning and optimization.
LOCATION:ME C2 405 https://plan.epfl.ch/?room==ME%20C2%20405
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR