HAMSI: A Parallel Incremental Optimization Algorithm Using Quadratic Approximations for Solving Partially Separable Problems

  • Kamer Kaya, Figen Oztoprak, +5 authors M. Kaan Ozturk
  • Published 2015

Abstract

We propose HAMSI, a provably convergent incremental algorithm for solving large-scale partially separable optimization problems that frequently emerge in machine learning and inferential statistics. The algorithm is based on a local quadratic approximation and hence allows incorporating a second order curvature information to speed-up the convergence… (More)

Topics

Cite this paper

@inproceedings{Kaya2015HAMSIAP, title={HAMSI: A Parallel Incremental Optimization Algorithm Using Quadratic Approximations for Solving Partially Separable Problems}, author={Kamer Kaya and Figen Oztoprak and cS. .Ilker Birbil and A. Taylan Cemgil and Umut cSimcsekli and Nurdan Kuru and Hazal Koptagel and M. Kaan Ozturk}, year={2015} }