##
Event Date

Speaker: Krishna Balasubramanian (UC Davis, Stat)

Title:** ****Normal Approximations for Stochastic Iterative Estimators (and Martingales)**

Abstract: Asymptotic normality of the maximum likelihood estimator (mle) is one of the foundational results of mathematical statistics characterizing the fluctuations of mle. But it suffers from two drawbacks: (i) it is asymptotic and (ii) it is established for the maximum likelihood estimator (i.e., argmin of negative log-likelihood function) which often can't be computed efficiently. Indeed, in practice the efficiently computable estimator is typically a stochastic iterative estimator/algorithm, which is run for a finite number of steps. The focus of this talk will be on establishing non-asymptotic normal approximation rates for such stochastic iterative estimators.

The first result of this talk is on establishing non-asymptotic normal approximation rates for stochastic gradient descent (SGD), arguably the most widely used stochastic iterative estimator, for locally strongly-convex (but globally potentially nonconvex) M-estimation problems. This result could be clubbed with existing bootstrap techniques to obtain non-asymptotically valid confidence sets for parameter estimation via the SGD estimator. The second result of this talk is on establishing non-asymptotic normal approximation rates for Euler discretizations of Itô diffusions (a special case of this estimator is the stochastic gradient Langevin Monte Carlo, widely used by the Bayesian community), which is a stochastic iterative estimator used for posterior expectation computation or numerical integration. This result could potentially be clubbed with (yet to be developed) bootstrap techniques to obtain non-asymptotically valid Frequentist-style confidence intervals for prediction within the Bayesian framework (if you are a statistician) or non-asymptotically valid confidence intervals for numerical integration in general (if you are an applied mathematician). The first and second result is proved by establishing non-asymptotic normal approximation rates for (multivariate) martingales by combining Stein's method with Lindeberg method and Skorokhod embedding method, respectively. These probabilistic results about martingales are of independent interest.

**Roundtable Discussion will start immediately after this talk. **The topics include:** ***Challenges in uncertainty quantification in data science; UCD4IDS activities in Winter 2020.*

So, please attend both the seminar and the roundtable discussion!