Distributed On-line Optimization and Learning for Nonconvex Big Data Analytics over Dynamic Networks
Abstract
Proposal Summary/Abstract This proposal aspires to i) devise parsimonious matrix and tensor inference models capturing a wide range of learning and processing tasks, which is a crucial step to obtain new insights in many defense applications; and then ii) to offer novel architectures and provable algorithms enabling nonconvex streaming analytics using parallel processors (when data are stored in a shared memory/storage) or distributed in-network processing (where data are distributed over time-varying networks).
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 03, 2016
- Source ID
- N000141612244
Entities
People
- Gesualdo Scutari
Organizations
- Office of Naval Research
- United States Navy
- University of Virginia