Jose Blanchet

Bio:

Jose Blanchet is a Professor of Management Science and Engineering (MS&E) at Stanford. Prior to joining MS&E, he was a Professor at the Departments of IEOR and Statistics at Columbia and before that a Professor of Statistics at Harvard. Jose is a recipient of the 2010 Erlang Prize (given every two years by INFORMS to young probabilists for outstanding contributions in applied probability) and has won several best publication awards in areas such as applied probability, simulation, operations management, and revenue management – including the biennial Best Publication Award given by the INFORMS Applied Probability Society biennially and the Outstanding Simulation Publication Award given to a single publication every year by the INFORMS Simulation Society). He also received a Presidential Early Career Award for Scientists and Engineers in 2010. He currently leads a Department of Defense sponsored Multi-University Research Initiative involving teams from Duke, Harvard, Maryland, MIT and Stanford on rare event analysis. He was the President of the INFORMS Applied Probability Society during 2020-2022 and has participated in various INFORMS committees. He is an area editor of Stochastic Models in Mathematics of Operations Research. He has served on the editorial board of Advances in Applied Probability, Bernoulli, Extremes, Insurance: Mathematics and Economics, Journal of Applied Probability, Queueing Systems: Theory and Applications, and Stochastic Systems, among others.

Title of talk:

Stochastic Optimization Algorithms with Heavy Tailed Input

Abstract of talk:

We first explain why statistical analysis of stochastic optimization algorithms with heavy-tailed input arises naturally in applications. In fact, we will argue that models that assume infinite variance gradient estimators in stochastic gradient descent are appropriate depending on easy-to-monitor features of historical data and on the spatial and temporal scales over which the algorithm will be deployed (even if models have finite variance in theory). We will then discuss inference tools that can be applied to monitor convergence of stochastic optimization algorithms based on several asymptotic statistics. The results that we will present include the first weak convergence analysis of stochastic gradient descent with infinite variance, extending results which assume finite variance or homogeneous and additive gradient noise. Based on joint work with Aleks Mijatovic, Wenhao Yang.