New Theory and Algorithms for Scalable Data Fusion
Abstract
The research performed under this grant served to address the modeling, algorithmic and theoretical challenges associated with problems of large-scale data fusion. Significant research accomplishments included: (a) the development of message passing algorithms for distributed optimization and inference; (b) the formulation and analysis of convex relaxations for estimating low-rank matrices from data; (c) the development of non-parametric methods for solving high-dimensional prediction problems; and (d) the analysis and implementation of methods for graphical model selection.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 14, 2013
- Accession Number
- ADA588861
Entities
People
- Martin J. Wainwright
Organizations
- University of California, Berkeley