Statistical Inference over Networks: Decentralized Optimization Meets High-dimensional Statistics

There is growing interest in solving large-scale statistical machine learning problems over decentralized networks, where data are distributed across the nodes of the network and no centralized coordination is present (we termed these systems as “mesh” networks). Inference from massive datasets poses a fundamental challenge at the nexus of the computational and statistical sciences: ensuring the quality of statistical inference when computational resources, like time and communication, are constrained. While statistical-computation tradeoffs have been largely explored in the centralized setting, our understanding over mesh networks is limited: (i) distributed schemes, designed and performing well in the classical low-dimensional regime, can break down in the high-dimensional case; and (ii) existing convergence studies may fail to predict algorithmic behaviors, with some findings directly contradicted by empirical tests. This is mainly due to the fact that the majority of distributed algorithms have been designed and studied only from the optimization perspective, lacking the statistical dimension. This talk will discuss some vignettes from high-dimensional statistical inference suggesting new analyses (and designs) aiming at bringing statistical thinking in distributed optimization.

Co-sponsored by: Fairleigh Dickinson University

Speaker(s): Dr. Gesualdo Scutari ,

Agenda:
There is growing interest in solving large-scale statistical machine learning problems over decentralized networks, where data are distributed across the nodes of the network and no centralized coordination is present (we termed these systems as “mesh” networks). Inference from massive datasets poses a fundamental challenge at the nexus of the computational and statistical sciences: ensuring the quality of statistical inference when computational resources, like time and communication, are constrained. While statistical-computation tradeoffs have been largely explored in the centralized setting, our understanding over mesh networks is limited: (i) distributed schemes, designed and performing well in the classical low-dimensional regime, can break down in the high-dimensional case; and (ii) existing convergence studies may fail to predict algorithmic behaviors, with some findings directly contradicted by empirical tests. This is mainly due to the fact that the majority of distributed algorithms have been designed and studied only from the optimization perspective, lacking the statistical dimension. This talk will discuss some vignettes from high-dimensional statistical inference suggesting new analyses (and designs) aiming at bringing statistical thinking in distributed optimization.

Virtual: https://events.vtools.ieee.org/m/380601