Communication remains the most significant bottleneck in the performance of distributed optimization algorithms for large-scale machine learning. In this paper, we propose a communication-efficient framework, CoCoA, that uses local computation in a primal-dual setting to dramatically reduce the amount of necessary communication. We provide a strong convergence rate analysis for this class of algorithms, as well as experiments on real-world distributed datasets with implementations in Spark. In our experiments, we find that as compared to state-of-the-art mini-batch versions of SGD and SDCA algorithms, CoCoA converges to the same .001-accurate solution quality on average 25x as quickly.
@article{arxiv.1409.1458,
title = {Communication-Efficient Distributed Dual Coordinate Ascent},
author = {Martin Jaggi and Virginia Smith and Martin Takáč and Jonathan Terhorst and Sanjay Krishnan and Thomas Hofmann and Michael I. Jordan},
journal= {arXiv preprint arXiv:1409.1458},
year = {2014}
}
Comments
NIPS 2014 version, including proofs. Published in Advances in Neural Information Processing Systems 27 (NIPS 2014)