CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: Pathwise optimization for adaptive bridge-type estimators and its application to SDEs Authors:  Alessandro De Gregorio - University of Rome La Sapienza (Italy) [presenting]
Francesco Iafrate - University of Rome La Sapienza (Italy)
Abstract: The focus is on the bridge-type estimators arising from optimization problems with multiple adaptive $l^q$-penalties. By resorting to some tools arising from the nonconvex optimization theory, we introduce algorithms for computing the full solution path for the introduced estimators, for any possible value of the penalization parameter. We highlight that, up to our knowledge, this is the first attempt to introduce computational efficient methods in the bridge estimation setting. Furthermore, we discuss some applications of our approach to stochastic differential equations.