Unwrapping ADMM: Efficient Distributed Computing via Transpose Reduction

Abstract

Recent approaches to distributed model fitting rely heavily on consensus ADMM, where each node solves small sub-problems using only local data. We propose iterative methods that solve global sub-problems over an entire distributed dataset. This is possible using transpose reduction strategies that allow a single node to solve least-squares over massive datasets with outputting all the data in one place. This results in simple iterative methods that avoid the expensive inner loops required for consensus methods. We analyze the convergence rates of the proposed schemes and demonstrate the efficiency of this approach by fitting linear classifiers and sparse linear models to large datasets using a distributed implementation with up to 20,000 cores in far less time than previous approaches.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 11, 2016
Accession Number
AD1014937

Entities

People

  • Gavin J. Taylor
  • Kawika Barabin
  • Kent Sayre
  • Thomas Goldstein

Organizations

  • United States Naval Academy

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Computations
  • Convergence
  • Data Sets
  • Dimensionality Reduction
  • Distributed Computing
  • Image Processing
  • Information Science
  • Iterations
  • Machine Learning
  • Materials
  • Optimization
  • Statistics
  • Supervised Machine Learning
  • United States Naval Academy

Fields of Study

  • Computer science

Readers

  • Approximation Theory.
  • Distributed Systems and Data Platform Development
  • Operations Research