AIDE: Fast and Communication Efficient Distributed Optimization

TL;DR

This paper introduces AIDE, an accelerated, communication-efficient distributed optimization algorithm that matches theoretical lower bounds and outperforms existing methods in machine learning tasks.

Contribution

The paper proposes AIDE, an accelerated variant of DANE, achieving optimal communication complexity and practical efficiency with a first-order oracle.

Findings

AIDE matches the communication lower bounds for distributed optimization.

AIDE outperforms existing communication-efficient algorithms in machine learning scenarios.

The inexact DANE variant maintains theoretical guarantees while improving practical robustness.

Abstract

In this paper, we present two new communication-efficient methods for distributed minimization of an average of functions. The first algorithm is an inexact variant of the DANE algorithm that allows any local algorithm to return an approximate solution to a local subproblem. We show that such a strategy does not affect the theoretical guarantees of DANE significantly. In fact, our approach can be viewed as a robustification strategy since the method is substantially better behaved than DANE on data partition arising in practice. It is well known that DANE algorithm does not match the communication complexity lower bounds. To bridge this gap, we propose an accelerated variant of the first method, called AIDE, that not only matches the communication lower bounds but can also be implemented using a purely first-order oracle. Our empirical results show that AIDE is superior to other communication efficient algorithms in settings that naturally arise in machine learning applications.