B1305
Title: Survey schemes for stochastic gradient descent with applications to M-estimation
Authors: Emilie Chautru - MINES ParisTech (France) [presenting]
Stephan Clemencon - Telecom ParisTech (France)
Patrice Bertail - Université Paris-Nanterre and TelecomParisTech (France)
Guillaume Papa - Telecom Paristech (France)
Abstract: The main purpose is to investigate the impact of survey sampling with unequal inclusion probabilities on stochastic gradient descent-based M-estimation methods in large-scale statistical and machine-learning problems. Precisely, when possible, we propose to take advantage of some auxiliary information to increase asymptotic accuracy of the estimation. The method is discussed in the specific context of big data.
