Talk of Prof. Cristian Rojas

May 7, 2024

--- Title: The Two-Stage Method: A Computationally Efficient and Private Approach to Statistical Learning

Time: May 7, 2024
Download as iCal:

Prof. Cristian Rojas
Division of Decision and Control Systems
KTH Royal Institute of Technology
Stockholm, Sweden   

 

Tuesday 2024-05-07 4 p.m.
IST Seminar Room 2.255 - Pfaffenwaldring 9 - Campus Stuttgart-Vaihingen

 

Abstract

The Two-Stage (TS) method is a computationally attractive approach to parameter estimation, which only requires the availability of a parametric model capable of simulating samples. Based on several values of the parameter to be estimated, and simulated samples for those values, TS can construct an efficient estimator using state-of-the-art supervised learning tools such as deep neural networks or gradient boosting.
In this talk, we will introduce the TS method from a statistical decision theory perspective (leading to Bayes and minimax versions), extend it to data-driven control applications, and, if time permits, discuss how to impose privacy constraints on this method.

 

Biographical Information

Cristian R. Rojas was born in 1980. He received the M.S. degree in electronics engineering from Universidad Técnica Federico Santa María, Valparaíso, Chile, in 2004, and the Ph.D. degree in electrical engineering at The University of Newcastle, NSW, Australia, in 2008. Since October 2008, he has been with the Royal Institute of Technology, Stockholm, Sweden, where he is currently Professor of Automatic Control at the School of Electrical Engineering and Computer Science. His research interests lie in system identification, signal processing and machine learning. Prof. Rojas is a Senior Member of IEEE, of the IEEE Technical Committee on System Identification and Adaptive Processing (since 2013), and of the IFAC Technical Committee TC1.1. on Modelling, Identification, and Signal Processing (since 2013). He has been Associate Editor for the IFAC journal Automatica and for the IEEE Control Systems Letters (L-CSS).

 
To the top of the page